korg/linux
0
0
Fork 0
mirror of https://mirrors.bfsu.edu.cn/git/linux.git synced 2024-12-29 05:55:02 +08:00
Code Issues Packages Projects Releases Wiki Activity
5a0e81f6f4
linux/drivers/infiniband/ulp/ipoib/ipoib_multicast.c

1032 lines
28 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/*
* Copyright (c) 2004, 2005 Topspin Communications. All rights reserved.
[PATCH] IB: Add copyright notices Make some lawyers happy and add copyright notices for people who forgot to include them when they actually touched the code. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-08-11 14:03:10 +08:00
* Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved.
* Copyright (c) 2004 Voltaire, Inc. All rights reserved.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
* General Public License (GPL) Version 2, available from the file
* COPYING in the main directory of this source tree, or the
* OpenIB.org BSD license below:
*
* Redistribution and use in source and binary forms, with or
* without modification, are permitted provided that the following
* conditions are met:
*
* - Redistributions of source code must retain the above
* copyright notice, this list of conditions and the following
* disclaimer.
*
* - Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials
* provided with the distribution.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#include <linux/skbuff.h>
#include <linux/rtnetlink.h>
infiniband: add moduleparam.h to drivers/infiniband as required These files were getting the moduleparam infrastructure from the implicit presence of module.h being everywhere, but that is going away soon. Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-08-31 00:32:52 +08:00
#include <linux/moduleparam.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include <linux/ip.h>
#include <linux/in.h>
#include <linux/igmp.h>
#include <linux/inetdevice.h>
#include <linux/delay.h>
#include <linux/completion.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
[INET_SOCK]: Move struct inet_sock & helper functions to net/inet_sock.h To help in reducing the number of include dependencies, several files were touched as they were getting needed headers indirectly for stuff they use. Thanks also to Alan Menegotto for pointing out that net/dccp/proto.c had linux/dccp.h include twice. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-27 12:43:12 +08:00
#include <net/dst.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include "ipoib.h"
#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
static int mcast_debug_level;
module_param(mcast_debug_level, int, 0644);
MODULE_PARM_DESC(mcast_debug_level,
"Enable multicast debug tracing if > 0");
#endif
struct ipoib_mcast_iter {
struct net_device *dev;
union ib_gid mgid;
unsigned long created;
unsigned int queuelen;
unsigned int complete;
unsigned int send_only;
};
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
* This should be called with the priv->lock held
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
*/
static void __ipoib_mcast_schedule_join_thread(struct ipoib_dev_priv *priv,
struct ipoib_mcast *mcast,
bool delay)
{
IB/ipoib: Remove IPOIB_MCAST_RUN bit After Doug Ledford's changes there is no need in that bit, it's semantic becomes subset of the IPOIB_FLAG_OPER_UP bit. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-04-02 18:39:04 +08:00
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
return;
/*
* We will be scheduling *something*, so cancel whatever is
* currently scheduled first
*/
cancel_delayed_work(&priv->mcast_task);
if (mcast && delay) {
/*
* We had a failure and want to schedule a retry later
*/
mcast->backoff *= 2;
if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
mcast->delay_until = jiffies + (mcast->backoff * HZ);
/*
* Mark this mcast for its delay, but restart the
* task immediately. The join task will make sure to
* clear out all entries without delays, and then
* schedule itself to run again when the earliest
* delay expires
*/
queue_delayed_work(priv->wq, &priv->mcast_task, 0);
} else if (delay) {
/*
* Special case of retrying after a failure to
* allocate the broadcast multicast group, wait
* 1 second and try again
*/
queue_delayed_work(priv->wq, &priv->mcast_task, HZ);
} else
queue_delayed_work(priv->wq, &priv->mcast_task, 0);
}
IB/IPoIB: factor out common multicast list removal code Code cleanup to remove multicast specific code from ipoib_main.c The removal of a list of multicast groups occurs in three places. Create a new function ipoib_mcast_remove_list(). Use this new function in ipoib_main.c too. That in turn allows the dropping of two functions that were exported from ipoib_multicast.c for expiration of mc groups. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-12-21 22:42:53 +08:00
static void ipoib_mcast_free(struct ipoib_mcast *mcast)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct net_device *dev = mcast->dev;
IPoIB: Lock accesses to multicast packet queues Avoid corrupting mcast->pkt_queue by serializing access with priv->tx_lock. Also, update dropped packet statistics to count multicast packets removed from pkt_queue as dropped. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-18 04:19:40 +08:00
int tx_dropped = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(netdev_priv(dev), "deleting multicast group %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IPoIB: Use a private hash table for path lookup in xmit path Dave Miller <davem@davemloft.net> provided a detailed description of why the way IPoIB is using neighbours for its own ipoib_neigh struct is buggy: Any time an ipoib_neigh is changed, a sequence like the following is made: spin_lock_irqsave(&priv->lock, flags); /* * It's safe to call ipoib_put_ah() inside * priv->lock here, because we know that * path->ah will always hold one more reference, * so ipoib_put_ah() will never do more than * decrement the ref count. */ if (neigh->ah) ipoib_put_ah(neigh->ah); list_del(&neigh->list); ipoib_neigh_free(dev, neigh); spin_unlock_irqrestore(&priv->lock, flags); ipoib_path_lookup(skb, n, dev); This doesn't work, because you're leaving a stale pointer to the freed up ipoib_neigh in the special neigh->ha pointer cookie. Yes, it even fails with all the locking done to protect _changes_ to *ipoib_neigh(n), and with the code in ipoib_neigh_free() that NULLs out the pointer. The core issue is that read side calls to *to_ipoib_neigh(n) are not being synchronized at all, they are performed without any locking. So whether we hold the lock or not when making changes to *ipoib_neigh(n) you still can have threads see references to freed up ipoib_neigh objects. cpu 1 cpu 2 n = *ipoib_neigh() *ipoib_neigh() = NULL kfree(n) n->foo == OOPS [..] Perhaps the ipoib code can have a private path database it manages entirely itself, which holds all the necessary information and is looked up by some generic key which is available easily at transmit time and does not involve generic neighbour entries. See <http://marc.info/?l=linux-rdma&m=132812793105624&w=2> and <http://marc.info/?l=linux-rdma&w=2&r=1&s=allows+references+to+freed+memory&q=b> for the full discussion. This patch aims to solve the race conditions found in the IPoIB driver. The patch removes the connection between the core networking neighbour structure and the ipoib_neigh structure. In addition to avoiding the race described above, it allows us to handle SKBs carrying IP packets that don't have any associated neighbour. We add an ipoib_neigh hash table with N buckets where the key is the destination hardware address. The ipoib_neigh is fetched from the hash table and instead of the stashed location in the neighbour structure. The hash table uses both RCU and reference counting to guarantee that no ipoib_neigh instance is ever deleted while in use. Fetching the ipoib_neigh structure instance from the hash also makes the special code in ipoib_start_xmit that handles remote and local bonding failover redundant. Aged ipoib_neigh instances are deleted by a garbage collection task that runs every M seconds and deletes every ipoib_neigh instance that was idle for at least 2*M seconds. The deletion is safe since the ipoib_neigh instances are protected using RCU and reference count mechanisms. The number of buckets (N) and frequency of running the GC thread (M), are taken from the exported arb_tbl. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-25 01:05:22 +08:00
/* remove all neigh connected to this mcast */
ipoib_del_neighs_by_gid(dev, mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (mcast->ah)
ipoib_put_ah(mcast->ah);
IPoIB: Lock accesses to multicast packet queues Avoid corrupting mcast->pkt_queue by serializing access with priv->tx_lock. Also, update dropped packet statistics to count multicast packets removed from pkt_queue as dropped. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-18 04:19:40 +08:00
while (!skb_queue_empty(&mcast->pkt_queue)) {
++tx_dropped;
[IPoIB] no need to set skb->dev right before freeing skb For cut-and-paste reasons, the IPoIB driver was setting skb->dev right before calling dev_kfree_skb_any(). Get rid of this. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-08 02:49:38 +08:00
dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue));
IPoIB: Lock accesses to multicast packet queues Avoid corrupting mcast->pkt_queue by serializing access with priv->tx_lock. Also, update dropped packet statistics to count multicast packets removed from pkt_queue as dropped. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-18 04:19:40 +08:00
}
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_lock_bh(dev);
[IPoIB]: Convert to netdevice internal stats Use the stats member of struct netdevice in IPoIB, so we can save memory by deleting the stats member of struct ipoib_dev_priv, and save code by deleting ipoib_get_stats(). Signed-off-by: Roland Dreier <rolandd@cisco.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-09-29 06:33:51 +08:00
dev->stats.tx_dropped += tx_dropped;
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_unlock_bh(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
kfree(mcast);
}
static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev,
int can_sleep)
{
struct ipoib_mcast *mcast;
[IB] kzalloc() conversions Replace kmalloc()+memset(,0,) with kzalloc(), for a net savings of 35 source lines and about 500 bytes of text. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-02 23:23:14 +08:00
mcast = kzalloc(sizeof *mcast, can_sleep ? GFP_KERNEL : GFP_ATOMIC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!mcast)
return NULL;
mcast->dev = dev;
mcast->created = jiffies;
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
mcast->delay_until = jiffies;
[PATCH] IPoIB: Fix SA client retransmission strategy We got a little mixed up with what the backoff member holds in the IPoIB multicast group structure: sometimes it was used as a number of seconds, and sometimes it was used as a number of jiffies. Fix the code so that backoff is always in seconds. Signed-off-by: Hal Rosenstock <halr@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-09-19 04:47:53 +08:00
mcast->backoff = 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
INIT_LIST_HEAD(&mcast->list);
INIT_LIST_HEAD(&mcast->neigh_list);
skb_queue_head_init(&mcast->pkt_queue);
return mcast;
}
IB/ipoib: Expire sendonly multicast joins On neighbor expiration, check to see if the neighbor was actually a sendonly multicast join, and if so, leave the multicast group as we expire the neighbor. Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-25 01:00:05 +08:00
struct ipoib_mcast *__ipoib_mcast_find(struct net_device *dev, void *mgid)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct rb_node *n = priv->multicast_tree.rb_node;
while (n) {
struct ipoib_mcast *mcast;
int ret;
mcast = rb_entry(n, struct ipoib_mcast, rb_node);
IPoIB: Fix kernel unaligned access on ia64 Fix misaligned access faults on ia64: never cast a misaligned neighbour->ha + 4 pointer to union ib_gid type; pass a void * pointer instead. The memcpy was being optimized to use full word accesses because the compiler thought that union ib_gid is always aligned. The cast in IPOIB_GID_ARG is safe, since it is fixed to access each byte separately. Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-05-30 00:14:05 +08:00
ret = memcmp(mgid, mcast->mcmember.mgid.raw,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sizeof (union ib_gid));
if (ret < 0)
n = n->rb_left;
else if (ret > 0)
n = n->rb_right;
else
return mcast;
}
return NULL;
}
static int __ipoib_mcast_add(struct net_device *dev, struct ipoib_mcast *mcast)
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct rb_node **n = &priv->multicast_tree.rb_node, *pn = NULL;
while (*n) {
struct ipoib_mcast *tmcast;
int ret;
pn = *n;
tmcast = rb_entry(pn, struct ipoib_mcast, rb_node);
ret = memcmp(mcast->mcmember.mgid.raw, tmcast->mcmember.mgid.raw,
sizeof (union ib_gid));
if (ret < 0)
n = &pn->rb_left;
else if (ret > 0)
n = &pn->rb_right;
else
return -EEXIST;
}
rb_link_node(&mcast->rb_node, pn, n);
rb_insert_color(&mcast->rb_node, &priv->multicast_tree);
return 0;
}
static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast,
struct ib_sa_mcmember_rec *mcmember)
{
struct net_device *dev = mcast->dev;
struct ipoib_dev_priv *priv = netdev_priv(dev);
IPoIB: Close race in setting mcast->ah ipoib_mcast_send() tests mcast->ah twice. If this value is changed between these two points, we leak an skb. However, ipoib_mcast_join_finish() sets mcast->ah with no locking, so it could race against ipoib_mcast_send(). As a solution, take priv->lock around assignment to mcast->ah thus making sure ipoib_mcast_send() (which also takes priv->lock) is not in flight. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-28 12:47:43 +08:00
struct ipoib_ah *ah;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
int ret;
IPoIB: Only set Q_Key once: after joining broadcast group The current code will set the Q_Key for any join of a non-sendonly multicast group. The operation involves a modify QP operation, which is fairly heavyweight, and is only really required after the join of the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach() to control when the Q_Key is set. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-07-15 14:48:50 +08:00
int set_qkey = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
mcast->mcmember = *mcmember;
IPoIB: Fix use-after-free of multicast object Fix a crash in ipoib_mcast_join_task(). (with help from Or Gerlitz) Commit c8c2afe360b7 ("IPoIB: Use rtnl lock/unlock when changing device flags") added a call to rtnl_lock() in ipoib_mcast_join_task(), which is run from the ipoib_workqueue, and hence the workqueue can't be flushed from the context of ipoib_stop(). In the current code, ipoib_stop() (which doesn't flush the workqueue) calls ipoib_mcast_dev_flush(), which goes and deletes all the multicast entries. This takes place without any synchronization with a possible running instance of ipoib_mcast_join_task() for the same ipoib device, leading to a crash due to NULL pointer dereference. Fix this by making sure that the workqueue is flushed before ipoib_mcast_dev_flush() is called. To make that possible, we move the RTNL-lock wrapped code to ipoib_mcast_join_finish(). Signed-off-by: Patrick McHardy <kaber@trash.net> Cc: <stable@vger.kernel.org> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-08-30 15:01:30 +08:00
/* Set the multicast MTU and cached Q_Key before we attach if it's
* the broadcast group.
*/
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!memcmp(mcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
sizeof (union ib_gid))) {
IPoIB: Test for NULL broadcast object in ipiob_mcast_join_finish() We saw a kernel oops in our regression testing when a multicast "join finish" occurred just after the interface was -- this is <https://bugs.openfabrics.org/show_bug.cgi?id=1040>. The test randomly causes the HCA physical port to go down then up. The cause of this is that ipoib_mcast_join_finish() processing happen just after ipoib_mcast_dev_flush() was invoked (in which case the broadcast pointer is NULL). This patch tests for and handles the case where priv->broadcast is NULL. Cc: <stable@kernel.org> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-05-21 06:41:09 +08:00
spin_lock_irq(&priv->lock);
if (!priv->broadcast) {
spin_unlock_irq(&priv->lock);
return -EAGAIN;
}
IB/ipoib: Update broadcast record values after each successful join request Update the cached broadcast record in the priv object after every new join of this broadcast domain group. These values are needed for the port configuration (MTU size) and to all the new multicast (non-broadcast) join requests initial parameters. For example, SM starts with 2K MTU for all the fabric, and after that it restarts (or handover to new SM) with new port configuration of 4K MTU. Without using the new values, the driver will keep its old configuration of 2K and will not apply the new configuration of 4K. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-04-02 18:39:01 +08:00
/*update priv member according to the new mcast*/
priv->broadcast->mcmember.qkey = mcmember->qkey;
priv->broadcast->mcmember.mtu = mcmember->mtu;
priv->broadcast->mcmember.traffic_class = mcmember->traffic_class;
priv->broadcast->mcmember.rate = mcmember->rate;
priv->broadcast->mcmember.sl = mcmember->sl;
priv->broadcast->mcmember.flow_label = mcmember->flow_label;
priv->broadcast->mcmember.hop_limit = mcmember->hop_limit;
/* assume if the admin and the mcast are the same both can be changed */
if (priv->mcast_mtu == priv->admin_mtu)
priv->admin_mtu =
priv->mcast_mtu =
IPOIB_UD_MTU(ib_mtu_enum_to_int(priv->broadcast->mcmember.mtu));
else
priv->mcast_mtu =
IPOIB_UD_MTU(ib_mtu_enum_to_int(priv->broadcast->mcmember.mtu));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey);
IPoIB: Test for NULL broadcast object in ipiob_mcast_join_finish() We saw a kernel oops in our regression testing when a multicast "join finish" occurred just after the interface was -- this is <https://bugs.openfabrics.org/show_bug.cgi?id=1040>. The test randomly causes the HCA physical port to go down then up. The cause of this is that ipoib_mcast_join_finish() processing happen just after ipoib_mcast_dev_flush() was invoked (in which case the broadcast pointer is NULL). This patch tests for and handles the case where priv->broadcast is NULL. Cc: <stable@kernel.org> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-05-21 06:41:09 +08:00
spin_unlock_irq(&priv->lock);
IB: split struct ib_send_wr This patch split up struct ib_send_wr so that all non-trivial verbs use their own structure which embedds struct ib_send_wr. This dramaticly shrinks the size of a WR for most common operations: sizeof(struct ib_send_wr) (old): 96 sizeof(struct ib_send_wr): 48 sizeof(struct ib_rdma_wr): 64 sizeof(struct ib_atomic_wr): 96 sizeof(struct ib_ud_wr): 88 sizeof(struct ib_fast_reg_wr): 88 sizeof(struct ib_bind_mw_wr): 96 sizeof(struct ib_sig_handover_wr): 80 And with Sagi's pending MR rework the fast registration WR will also be down to a reasonable size: sizeof(struct ib_fastreg_wr): 64 Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com> [srp, srpt] Reviewed-by: Chuck Lever <chuck.lever@oracle.com> [sunrpc] Tested-by: Haggai Eran <haggaie@mellanox.com> Tested-by: Sagi Grimberg <sagig@mellanox.com> Tested-by: Steve Wise <swise@opengridcomputing.com>
2015-10-08 16:16:33 +08:00
priv->tx_wr.remote_qkey = priv->qkey;
IPoIB: Only set Q_Key once: after joining broadcast group The current code will set the Q_Key for any join of a non-sendonly multicast group. The operation involves a modify QP operation, which is fairly heavyweight, and is only really required after the join of the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach() to control when the Q_Key is set. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-07-15 14:48:50 +08:00
set_qkey = 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
if (test_and_set_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_warn(priv, "multicast group %pI6 already attached\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
ret = ipoib_mcast_attach(dev, be16_to_cpu(mcast->mcmember.mlid),
IPoIB: Only set Q_Key once: after joining broadcast group The current code will set the Q_Key for any join of a non-sendonly multicast group. The operation involves a modify QP operation, which is fairly heavyweight, and is only really required after the join of the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach() to control when the Q_Key is set. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-07-15 14:48:50 +08:00
&mcast->mcmember.mgid, set_qkey);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (ret < 0) {
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_warn(priv, "couldn't attach QP to multicast group %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags);
return ret;
}
}
{
struct ib_ah_attr av = {
.dlid = be16_to_cpu(mcast->mcmember.mlid),
.port_num = priv->port,
.sl = mcast->mcmember.sl,
.ah_flags = IB_AH_GRH,
IB: simplify static rate encoding Push translation of static rate to HCA format into low-level drivers, where it belongs. For static rate encoding, use encoding of rate field from IB standard PathRecord, with addition of value 0, for backwards compatibility with current usage. The changes are: - Add enum ib_rate to midlayer includes. - Get rid of static rate translation in IPoIB; just use static rate directly from Path and MulticastGroup records. - Update mthca driver to translate absolute static rate into the format used by hardware. This also fixes mthca's static rate handling for HCAs that are capable of 4X DDR. Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-04-11 00:43:47 +08:00
.static_rate = mcast->mcmember.rate,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
.grh = {
.flow_label = be32_to_cpu(mcast->mcmember.flow_label),
.hop_limit = mcast->mcmember.hop_limit,
.sgid_index = 0,
.traffic_class = mcast->mcmember.traffic_class
}
};
av.grh.dgid = mcast->mcmember.mgid;
IPoIB: Close race in setting mcast->ah ipoib_mcast_send() tests mcast->ah twice. If this value is changed between these two points, we leak an skb. However, ipoib_mcast_join_finish() sets mcast->ah with no locking, so it could race against ipoib_mcast_send(). As a solution, take priv->lock around assignment to mcast->ah thus making sure ipoib_mcast_send() (which also takes priv->lock) is not in flight. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-28 12:47:43 +08:00
ah = ipoib_create_ah(dev, priv->pd, &av);
IB/ipoib: Prevent hung task or softlockup processing multicast response This following can occur with ipoib when processing a multicast reponse: BUG: soft lockup - CPU#0 stuck for 67s! [ib_mad1:982] Modules linked in: ... CPU 0: Modules linked in: ... Pid: 982, comm: ib_mad1 Not tainted 2.6.32-131.0.15.el6.x86_64 #1 ProLiant DL160 G5 RIP: 0010:[<ffffffff814ddb27>] [<ffffffff814ddb27>] _spin_unlock_irqrestore+0x17/0x20 RSP: 0018:ffff8802119ed860 EFLAGS: 00000246 0000000000000004 RBX: ffff8802119ed860 RCX: 000000000000a299 RDX: ffff88021086c700 RSI: 0000000000000246 RDI: 0000000000000246 RBP: ffffffff8100bc8e R08: ffff880210ac229c R09: 0000000000000000 R10: ffff88021278aab8 R11: 0000000000000000 R12: ffff8802119ed860 R13: ffffffff8100be6e R14: 0000000000000001 R15: 0000000000000003 FS: 0000000000000000(0000) GS:ffff880028200000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 00000000006d4840 CR3: 0000000209aa5000 CR4: 00000000000406f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Call Trace: [<ffffffffa032c247>] ? ipoib_mcast_send+0x157/0x480 [ib_ipoib] [<ffffffff8100bc8e>] ? apic_timer_interrupt+0xe/0x20 [<ffffffff8100bc8e>] ? apic_timer_interrupt+0xe/0x20 [<ffffffffa03283d4>] ? ipoib_path_lookup+0x124/0x2d0 [ib_ipoib] [<ffffffffa03286fc>] ? ipoib_start_xmit+0x17c/0x430 [ib_ipoib] [<ffffffff8141e758>] ? dev_hard_start_xmit+0x2c8/0x3f0 [<ffffffff81439d0a>] ? sch_direct_xmit+0x15a/0x1c0 [<ffffffff81423098>] ? dev_queue_xmit+0x388/0x4d0 [<ffffffffa032d6b7>] ? ipoib_mcast_join_finish+0x2c7/0x510 [ib_ipoib] [<ffffffffa032dab8>] ? ipoib_mcast_sendonly_join_complete+0x1b8/0x1f0 [ib_ipoib] [<ffffffffa02a0946>] ? mcast_work_handler+0x1a6/0x710 [ib_sa] [<ffffffffa015f01e>] ? ib_send_mad+0xfe/0x3c0 [ib_mad] [<ffffffffa00f6c93>] ? ib_get_cached_lmc+0xa3/0xb0 [ib_core] [<ffffffffa02a0f9b>] ? join_handler+0xeb/0x200 [ib_sa] [<ffffffffa029e4fc>] ? ib_sa_mcmember_rec_callback+0x5c/0xa0 [ib_sa] [<ffffffffa029e79c>] ? recv_handler+0x3c/0x70 [ib_sa] [<ffffffffa01603a4>] ? ib_mad_completion_handler+0x844/0x9d0 [ib_mad] [<ffffffffa015fb60>] ? ib_mad_completion_handler+0x0/0x9d0 [ib_mad] [<ffffffff81088830>] ? worker_thread+0x170/0x2a0 [<ffffffff8108e160>] ? autoremove_wake_function+0x0/0x40 [<ffffffff810886c0>] ? worker_thread+0x0/0x2a0 [<ffffffff8108ddf6>] ? kthread+0x96/0xa0 [<ffffffff8100c1ca>] ? child_rip+0xa/0x20 Coinciding with stack trace is the following message: ib0: ib_address_create failed The code below in ipoib_mcast_join_finish() will note the above failure in the address handle but otherwise continue: ah = ipoib_create_ah(dev, priv->pd, &av); if (!ah) { ipoib_warn(priv, "ib_address_create failed\n"); } else { The while loop at the bottom of ipoib_mcast_join_finish() will attempt to send queued multicast packets in mcast->pkt_queue and eventually end up in ipoib_mcast_send(): if (!mcast->ah) { if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE) skb_queue_tail(&mcast->pkt_queue, skb); else { ++dev->stats.tx_dropped; dev_kfree_skb_any(skb); } My read is that the code will requeue the packet and return to the ipoib_mcast_join_finish() while loop and the stage is set for the "hung" task diagnostic as the while loop never sees a non-NULL ah, and will do nothing to resolve. There are GFP_ATOMIC allocates in the provider routines, so this is possible and should be dealt with. The test that induced the failure is associated with a host SM on the same server during a shutdown. This patch causes ipoib_mcast_join_finish() to exit with an error which will flush the queued mcast packets. Nothing is done to unwind the QP attached state so that subsequent sends from above will retry the join. Reviewed-by: Ram Vepa <ram.vepa@qlogic.com> Reviewed-by: Gary Leshner <gary.leshner@qlogic.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2011-11-21 21:43:54 +08:00
if (IS_ERR(ah)) {
ipoib_warn(priv, "ib_address_create failed %ld\n",
-PTR_ERR(ah));
/* use original error */
return PTR_ERR(ah);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
} else {
IB/ipoib: Fix oops with ipoib_debug_mcast set Need to set mcast->ah before debug code dereferences it. Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-07-24 15:42:00 +08:00
spin_lock_irq(&priv->lock);
mcast->ah = ah;
spin_unlock_irq(&priv->lock);
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "MGID %pI6 AV %p, LID 0x%04x, SL %d\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
mcast->ah->ah,
be16_to_cpu(mcast->mcmember.mlid),
mcast->mcmember.sl);
}
}
/* actually send any queued packets */
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_lock_bh(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
while (!skb_queue_empty(&mcast->pkt_queue)) {
struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue);
net: Abstract dst->neighbour accesses behind helpers. dst_{get,set}_neighbour() Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-18 14:09:49 +08:00
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_unlock_bh(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
skb->dev = dev;
if (dev_queue_xmit(skb))
ipoib_warn(priv, "dev_queue_xmit failed to requeue packet\n");
IPoIB: Stop lying about hard_header_len and use skb->cb to stash LL addresses Commit a0417fa3a18a ("net: Make qdisc_skb_cb upper size bound explicit.") made it possible for a netdev driver to use skb->cb between its header_ops.create method and its .ndo_start_xmit method. Use this in ipoib_hard_header() to stash away the LL address (GID + QPN), instead of the "ipoib_pseudoheader" hack. This allows IPoIB to stop lying about its hard_header_len, which will let us fix the L2 check for GRO. Signed-off-by: Roland Dreier <roland@purestorage.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-02-07 22:51:21 +08:00
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_lock_bh(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
netif_tx_unlock_bh(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop() Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with ipoib_stop(). We avoid it by scheduling the piece of code that takes the lock on ipoib_workqueue instead of executing it directly. This works because we only flush the ipoib_workqueue with the RTNL not held. The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down() which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(), which calls ipoib_mcast_leave(). The latter calls ib_sa_free_multicast(), and this waits until the multicast completion handler finishes. This handler is ipoib_mcast_join_complete(), which waits for the rtnl_lock(), which was already taken by ipoib_stop(). This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"). Signed-off-by: Yossi Etigin <yosefe@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-09-17 02:57:45 +08:00
void ipoib_mcast_carrier_on_task(struct work_struct *work)
{
struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv,
carrier_on_task);
IPoIB: Don't turn on carrier for a non-active port Multicast joins can succeed even if the IB port is down. This happens when the SM runs on the same port with the requesting port. However, IPoIB calls netif_carrier_on() when the join of the broadcast group succeeds, without caring about the state of the IB port. The result is an IPoIB interface in RUNNING state but without an active IB port to support it. If a bonding interface uses this IPoIB interface as a slave it might not detect that this slave is almost useless and failover functionality will be damaged. The fix checks the state of the IB port in the carrier_task before calling netif_carrier_on(). Adresses: https://bugs.openfabrics.org/show_bug.cgi?id=1726 Signed-off-by: Moni Shoua <monis@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-25 03:01:05 +08:00
struct ib_port_attr attr;
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop() Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with ipoib_stop(). We avoid it by scheduling the piece of code that takes the lock on ipoib_workqueue instead of executing it directly. This works because we only flush the ipoib_workqueue with the RTNL not held. The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down() which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(), which calls ipoib_mcast_leave(). The latter calls ib_sa_free_multicast(), and this waits until the multicast completion handler finishes. This handler is ipoib_mcast_join_complete(), which waits for the rtnl_lock(), which was already taken by ipoib_stop(). This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"). Signed-off-by: Yossi Etigin <yosefe@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-09-17 02:57:45 +08:00
IPoIB: Don't turn on carrier for a non-active port Multicast joins can succeed even if the IB port is down. This happens when the SM runs on the same port with the requesting port. However, IPoIB calls netif_carrier_on() when the join of the broadcast group succeeds, without caring about the state of the IB port. The result is an IPoIB interface in RUNNING state but without an active IB port to support it. If a bonding interface uses this IPoIB interface as a slave it might not detect that this slave is almost useless and failover functionality will be damaged. The fix checks the state of the IB port in the carrier_task before calling netif_carrier_on(). Adresses: https://bugs.openfabrics.org/show_bug.cgi?id=1726 Signed-off-by: Moni Shoua <monis@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-25 03:01:05 +08:00
if (ib_query_port(priv->ca, priv->port, &attr) ||
attr.state != IB_PORT_ACTIVE) {
ipoib_dbg(priv, "Keeping carrier off until IB port is active\n");
return;
}
IB/ipoib: Make the carrier_on_task race aware We blindly assume that we can just take the rtnl lock and that will prevent races with downing this interface. Unfortunately, that's not the case. In ipoib_mcast_stop_thread() we will call flush_workqueue() in an attempt to clear out all remaining instances of ipoib_join_task. But, since this task is put on the same workqueue as the join task, the flush_workqueue waits on this thread too. But this thread is deadlocked on the rtnl lock. The better thing here is to use trylock and loop on that until we either get the lock or we see that FLAG_OPER_UP has been cleared, in which case we don't need to do anything anyway and we just return. While investigating which flag should be used, FLAG_ADMIN_UP or FLAG_OPER_UP, it was determined that FLAG_OPER_UP was the more appropriate flag to use. However, there was a mix of these two flags in use in the existing code. So while we check for that flag here as part of this race fix, also cleanup the two places that had used the less appropriate flag for their tests. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:02 +08:00
/*
* Take rtnl_lock to avoid racing with ipoib_stop() and
* turning the carrier back on while a device is being
* removed. However, ipoib_stop() will attempt to flush
* the workqueue while holding the rtnl lock, so loop
* on trylock until either we get the lock or we see
* FLAG_OPER_UP go away as that signals that we are bailing
* and can safely ignore the carrier on work.
*/
while (!rtnl_trylock()) {
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
return;
else
msleep(20);
}
IB/ipoib: Consolidate rtnl_lock tasks in workqueue The ipoib_mcast_flush_dev routine is called with the rtnl_lock held and needs to keep it held. It also needs to call flush_workqueue() to flush out any outstanding work. In the past, we've had to try and make sure that we didn't flush out any outstanding join completions because they also wanted to grab rtnl_lock() and that would deadlock. It turns out that the only thing in the join completion handler that needs this lock can be safely moved to our carrier_on_task, thereby reducing the potential for the join completion code and the flush code to deadlock against each other. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:01 +08:00
if (!ipoib_cm_admin_enabled(priv->dev))
dev_set_mtu(priv->dev, min(priv->mcast_mtu, priv->admin_mtu));
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop() Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with ipoib_stop(). We avoid it by scheduling the piece of code that takes the lock on ipoib_workqueue instead of executing it directly. This works because we only flush the ipoib_workqueue with the RTNL not held. The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down() which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(), which calls ipoib_mcast_leave(). The latter calls ib_sa_free_multicast(), and this waits until the multicast completion handler finishes. This handler is ipoib_mcast_join_complete(), which waits for the rtnl_lock(), which was already taken by ipoib_stop(). This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"). Signed-off-by: Yossi Etigin <yosefe@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-09-17 02:57:45 +08:00
netif_carrier_on(priv->dev);
rtnl_unlock();
}
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
static int ipoib_mcast_join_complete(int status,
struct ib_sa_multicast *multicast)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
struct ipoib_mcast *mcast = multicast->context;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct net_device *dev = mcast->dev;
struct ipoib_dev_priv *priv = netdev_priv(dev);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
ipoib_dbg_mcast(priv, "%sjoin completion for %pI6 (status %d)\n",
test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) ?
"sendonly " : "",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw, status);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
/* We trap for port events ourselves. */
Revert "IPoIB: fix MCAST_FLAG_BUSY usage" This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:20 +08:00
if (status == -ENETRESET) {
status = 0;
IPoIB: Fix usage of uninitialized multicast objects The driver should avoid calling ib_sa_free_multicast on the mcast->mc object until it finishes its initialization state. Otherwise we can crash when ipoib_mcast_dev_flush() attempts to use the uninitialized multicast object. Instead, only call wait_for_completion() for multicast entries that started the join process, meaning that ib_sa_join_multicast() finished. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:51 +08:00
goto out;
Revert "IPoIB: fix MCAST_FLAG_BUSY usage" This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:20 +08:00
}
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
if (!status)
status = ipoib_mcast_join_finish(mcast, &multicast->rec);
if (!status) {
[PATCH] IPoIB: Fix SA client retransmission strategy We got a little mixed up with what the backoff member holds in the IPoIB multicast group structure: sometimes it was used as a number of seconds, and sometimes it was used as a number of jiffies. Fix the code so that backoff is always in seconds. Signed-off-by: Hal Rosenstock <halr@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-09-19 04:47:53 +08:00
mcast->backoff = 1;
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
mcast->delay_until = jiffies;
IPoIB: Turn on interface's carrier after broadcast group is joined Do netif_carrier_on() right after the IPv4 broadcast multicast group is joined, rather than waiting for all of the initial set of multicast group joins to finish. This allows at least IPv4 traffic to limp along on broken fabrics where not all multicast groups can be joined. Signed-off-by: Shirley Ma <xma@us.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-03-09 06:59:30 +08:00
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop() Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with ipoib_stop(). We avoid it by scheduling the piece of code that takes the lock on ipoib_workqueue instead of executing it directly. This works because we only flush the ipoib_workqueue with the RTNL not held. The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down() which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(), which calls ipoib_mcast_leave(). The latter calls ib_sa_free_multicast(), and this waits until the multicast completion handler finishes. This handler is ipoib_mcast_join_complete(), which waits for the rtnl_lock(), which was already taken by ipoib_stop(). This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"). Signed-off-by: Yossi Etigin <yosefe@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-09-17 02:57:45 +08:00
/*
IB/ipoib: Use dedicated workqueues per interface During my recent work on the rtnl lock deadlock in the IPoIB driver, I saw that even once I fixed the apparent races for a single device, as soon as that device had any children, new races popped up. It turns out that this is because no matter how well we protect against races on a single device, the fact that all devices use the same workqueue, and flush_workqueue() flushes *everything* from that workqueue means that we would also have to prevent all races between different devices (for instance, ipoib_mcast_restart_task on interface ib0 can race with ipoib_mcast_flush_dev on interface ib0.8002, resulting in a deadlock on the rtnl_lock). There are several possible solutions to this problem: Make carrier_on_task and mcast_restart_task try to take the rtnl for some set period of time and if they fail, then bail. This runs the real risk of dropping work on the floor, which can end up being its own separate kind of deadlock. Set some global flag in the driver that says some device is in the middle of going down, letting all tasks know to bail. Again, this can drop work on the floor. Or the method this patch attempts to use, which is when we bring an interface up, create a workqueue specifically for that interface, so that when we take it back down, we are flushing only those tasks associated with our interface. In addition, keep the global workqueue, but now limit it to only flush tasks. In this way, the flush tasks can always flush the device specific work queues without having deadlock issues. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:03 +08:00
* Defer carrier on work to priv->wq to avoid a
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
* deadlock on rtnl_lock here. Requeue our multicast
* work too, which will end up happening right after
* our carrier on task work and will allow us to
* send out all of the non-broadcast joins
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop() Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with ipoib_stop(). We avoid it by scheduling the piece of code that takes the lock on ipoib_workqueue instead of executing it directly. This works because we only flush the ipoib_workqueue with the RTNL not held. The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down() which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(), which calls ipoib_mcast_leave(). The latter calls ib_sa_free_multicast(), and this waits until the multicast completion handler finishes. This handler is ipoib_mcast_join_complete(), which waits for the rtnl_lock(), which was already taken by ipoib_stop(). This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"). Signed-off-by: Yossi Etigin <yosefe@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-09-17 02:57:45 +08:00
*/
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (mcast == priv->broadcast) {
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_lock_irq(&priv->lock);
IB/ipoib: Use dedicated workqueues per interface During my recent work on the rtnl lock deadlock in the IPoIB driver, I saw that even once I fixed the apparent races for a single device, as soon as that device had any children, new races popped up. It turns out that this is because no matter how well we protect against races on a single device, the fact that all devices use the same workqueue, and flush_workqueue() flushes *everything* from that workqueue means that we would also have to prevent all races between different devices (for instance, ipoib_mcast_restart_task on interface ib0 can race with ipoib_mcast_flush_dev on interface ib0.8002, resulting in a deadlock on the rtnl_lock). There are several possible solutions to this problem: Make carrier_on_task and mcast_restart_task try to take the rtnl for some set period of time and if they fail, then bail. This runs the real risk of dropping work on the floor, which can end up being its own separate kind of deadlock. Set some global flag in the driver that says some device is in the middle of going down, letting all tasks know to bail. Again, this can drop work on the floor. Or the method this patch attempts to use, which is when we bring an interface up, create a workqueue specifically for that interface, so that when we take it back down, we are flushing only those tasks associated with our interface. In addition, keep the global workqueue, but now limit it to only flush tasks. In this way, the flush tasks can always flush the device specific work queues without having deadlock issues. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:03 +08:00
queue_work(priv->wq, &priv->carrier_on_task);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
__ipoib_mcast_schedule_join_thread(priv, NULL, 0);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
goto out_locked;
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
}
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
} else {
IB/ipoib: Suppress warning for send only join failures We expect send only joins to fail, it just means there are no listeners for the group. The correct thing to do is silently drop the packet at source. Eg avahi will full join 224.0.0.251 which causes a send only IGMP packet to 224.0.0.22, and then a warning level kmessage like this: ib0: sendonly multicast join failed for ff12:401b:ffff:0000:0000:0000:0000:0016, status -22 If there is no IP router listening to IGMP. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-08-22 07:34:13 +08:00
bool silent_fail =
test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) &&
status == -EINVAL;
if (mcast->logcount < 20) {
if (status == -ETIMEDOUT || status == -EAGAIN ||
silent_fail) {
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
ipoib_dbg_mcast(priv, "%smulticast join failed for %pI6, status %d\n",
test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) ? "sendonly " : "",
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
mcast->mcmember.mgid.raw, status);
} else {
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
ipoib_warn(priv, "%smulticast join failed for %pI6, status %d\n",
test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) ? "sendonly " : "",
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
mcast->mcmember.mgid.raw, status);
}
IB/ipoib: Suppress warning for send only join failures We expect send only joins to fail, it just means there are no listeners for the group. The correct thing to do is silently drop the packet at source. Eg avahi will full join 224.0.0.251 which causes a send only IGMP packet to 224.0.0.22, and then a warning level kmessage like this: ib0: sendonly multicast join failed for ff12:401b:ffff:0000:0000:0000:0000:0016, status -22 If there is no IP router listening to IGMP. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-08-22 07:34:13 +08:00
if (!silent_fail)
mcast->logcount++;
Revert "IPoIB: fix MCAST_FLAG_BUSY usage" This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:20 +08:00
}
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) &&
mcast->backoff >= 2) {
/*
* We only retry sendonly joins once before we drop
* the packet and quit trying to deal with the
* group. However, we leave the group in the
* mcast list as an unjoined group. If we want to
* try joining again, we simply queue up a packet
* and restart the join thread. The empty queue
* is why the join thread ignores this group.
*/
mcast->backoff = 1;
netif_tx_lock_bh(dev);
while (!skb_queue_empty(&mcast->pkt_queue)) {
++dev->stats.tx_dropped;
dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue));
}
netif_tx_unlock_bh(dev);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
} else {
spin_lock_irq(&priv->lock);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
/* Requeue this join task with a backoff delay */
__ipoib_mcast_schedule_join_thread(priv, mcast, 1);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
goto out_locked;
}
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
}
Revert "IPoIB: fix MCAST_FLAG_BUSY usage" This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:20 +08:00
out:
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_lock_irq(&priv->lock);
out_locked:
/*
* Make sure to set mcast->mc before we clear the busy flag to avoid
* racing with code that checks for BUSY before checking mcast->mc
*/
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (status)
mcast->mc = NULL;
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
else
mcast->mc = multicast;
clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
spin_unlock_irq(&priv->lock);
Revert "IPoIB: fix MCAST_FLAG_BUSY usage" This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:20 +08:00
complete(&mcast->done);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
return status;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
struct ib_sa_multicast *multicast;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct ib_sa_mcmember_rec rec = {
.join_state = 1
};
ib_sa_comp_mask comp_mask;
int ret = 0;
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "joining MGID %pI6\n", mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rec.mgid = mcast->mcmember.mgid;
rec.port_gid = priv->local_gid;
[PATCH] IB: sparse endianness cleanup Fix sparse warnings. Use __be* where appropriate. Signed-off-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-08-14 12:05:57 +08:00
rec.pkey = cpu_to_be16(priv->pkey);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
comp_mask =
IB_SA_MCMEMBER_REC_MGID |
IB_SA_MCMEMBER_REC_PORT_GID |
IB_SA_MCMEMBER_REC_PKEY |
IB_SA_MCMEMBER_REC_JOIN_STATE;
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
if (mcast != priv->broadcast) {
/*
* RFC 4391:
* The MGID MUST use the same P_Key, Q_Key, SL, MTU,
* and HopLimit as those used in the broadcast-GID. The rest
* of attributes SHOULD follow the values used in the
* broadcast-GID as well.
*/
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
comp_mask |=
IPoIB: Create MCGs with all attributes required by RFC RFC 4391 ("Transmission of IP over InfiniBand (IPoIB)") says: If the IB multicast group does not already exist, one must be created first with the IPoIB link MTU. The MGID MUST use the same P_Key, Q_Key, SL, MTU, and HopLimit as those used in the broadcast-GID. The rest of attributes SHOULD follow the values used in the broadcast-GID as well. However, the current IPoIB driver is only setting the attributes required by the InfiniBand spec to create a multicast group, so in particular the MTU and HopLimit are not being set. Add these attributes when creating MCGs, and also set the Rate attribute, since IPoIB pays attention to that attribute as well. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-09-23 06:22:56 +08:00
IB_SA_MCMEMBER_REC_QKEY |
IB_SA_MCMEMBER_REC_MTU_SELECTOR |
IB_SA_MCMEMBER_REC_MTU |
IB_SA_MCMEMBER_REC_TRAFFIC_CLASS |
IB_SA_MCMEMBER_REC_RATE_SELECTOR |
IB_SA_MCMEMBER_REC_RATE |
IB_SA_MCMEMBER_REC_SL |
IB_SA_MCMEMBER_REC_FLOW_LABEL |
IB_SA_MCMEMBER_REC_HOP_LIMIT;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rec.qkey = priv->broadcast->mcmember.qkey;
IPoIB: Create MCGs with all attributes required by RFC RFC 4391 ("Transmission of IP over InfiniBand (IPoIB)") says: If the IB multicast group does not already exist, one must be created first with the IPoIB link MTU. The MGID MUST use the same P_Key, Q_Key, SL, MTU, and HopLimit as those used in the broadcast-GID. The rest of attributes SHOULD follow the values used in the broadcast-GID as well. However, the current IPoIB driver is only setting the attributes required by the InfiniBand spec to create a multicast group, so in particular the MTU and HopLimit are not being set. Add these attributes when creating MCGs, and also set the Rate attribute, since IPoIB pays attention to that attribute as well. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-09-23 06:22:56 +08:00
rec.mtu_selector = IB_SA_EQ;
rec.mtu = priv->broadcast->mcmember.mtu;
rec.traffic_class = priv->broadcast->mcmember.traffic_class;
rec.rate_selector = IB_SA_EQ;
rec.rate = priv->broadcast->mcmember.rate;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rec.sl = priv->broadcast->mcmember.sl;
rec.flow_label = priv->broadcast->mcmember.flow_label;
IPoIB: Create MCGs with all attributes required by RFC RFC 4391 ("Transmission of IP over InfiniBand (IPoIB)") says: If the IB multicast group does not already exist, one must be created first with the IPoIB link MTU. The MGID MUST use the same P_Key, Q_Key, SL, MTU, and HopLimit as those used in the broadcast-GID. The rest of attributes SHOULD follow the values used in the broadcast-GID as well. However, the current IPoIB driver is only setting the attributes required by the InfiniBand spec to create a multicast group, so in particular the MTU and HopLimit are not being set. Add these attributes when creating MCGs, and also set the Rate attribute, since IPoIB pays attention to that attribute as well. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-09-23 06:22:56 +08:00
rec.hop_limit = priv->broadcast->mcmember.hop_limit;
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
/*
IB/ipoib: Make sendonly multicast joins create the mcast group Since IPoIB should, as much as possible, emulate how multicast sends work on Ethernet for regular TCP/IP apps, there should be no requirement to subscribe to a multicast group before your sends are properly sent. However, due to the difference in how multicast is handled on InfiniBand, we must join the appropriate multicast group before we can send to it. Previously we tried not to trigger the auto-create feature of the subnet manager when doing this because we didn't have tracking of these sendonly groups and the auto-creation might never get undone. The previous patch added timing to these sendonly joins and allows us to leave them after a reasonable idle expiration time. So supply all of the information needed to auto-create group. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-26 02:35:01 +08:00
* Send-only IB Multicast joins do not work at the core
* IB layer yet, so we can't use them here. However,
* we are emulating an Ethernet multicast send, which
* does not require a multicast subscription and will
* still send properly. The most appropriate thing to
* do is to create the group if it doesn't exist as that
* most closely emulates the behavior, from a user space
* application perspecitive, of Ethernet multicast
* operation. For now, we do a full join, maybe later
* when the core IB layers support send only joins we
* will use them.
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
*/
IB/ipoib: Make sendonly multicast joins create the mcast group Since IPoIB should, as much as possible, emulate how multicast sends work on Ethernet for regular TCP/IP apps, there should be no requirement to subscribe to a multicast group before your sends are properly sent. However, due to the difference in how multicast is handled on InfiniBand, we must join the appropriate multicast group before we can send to it. Previously we tried not to trigger the auto-create feature of the subnet manager when doing this because we didn't have tracking of these sendonly groups and the auto-creation might never get undone. The previous patch added timing to these sendonly joins and allows us to leave them after a reasonable idle expiration time. So supply all of the information needed to auto-create group. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-26 02:35:01 +08:00
#if 0
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
rec.join_state = 4;
#endif
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
multicast = ib_sa_join_multicast(&ipoib_sa_client, priv->ca, priv->port,
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
&rec, comp_mask, GFP_KERNEL,
ipoib_mcast_join_complete, mcast);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
if (IS_ERR(multicast)) {
ret = PTR_ERR(multicast);
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
ipoib_warn(priv, "ib_sa_join_multicast failed, status %d\n", ret);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_lock_irq(&priv->lock);
/* Requeue this join task with a backoff delay */
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
__ipoib_mcast_schedule_join_thread(priv, mcast, 1);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
spin_unlock_irq(&priv->lock);
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
complete(&mcast->done);
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
WorkStruct: make allyesconfig Fix up for make allyesconfig. Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 22:57:56 +08:00
void ipoib_mcast_join_task(struct work_struct *work)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
WorkStruct: make allyesconfig Fix up for make allyesconfig. Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 22:57:56 +08:00
struct ipoib_dev_priv *priv =
container_of(work, struct ipoib_dev_priv, mcast_task.work);
struct net_device *dev = priv->dev;
IPoIB: Start multicast join process only on active ports The driver starts the mcast_join task whenever the netdev interface is UP without relation to the underlying IB port state. Until the port state is ACTIVE all the join requests are irrelevant, and the IB core returns -EINVAL. So the user will see errors such as: "multicast join failed for ff12:401b:... , status -22". Instead, have ipoib_mcast_join_task() return when the port is not active. It will be called again when the port state is changed and the low-level driver triggers the IB_EVENT_PORT_ACTIVE event or the IB_EVENT_CLIENT_REREGISTER event. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:53 +08:00
struct ib_port_attr port_attr;
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
unsigned long delay_until = 0;
struct ipoib_mcast *mcast = NULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: Remove IPOIB_MCAST_RUN bit After Doug Ledford's changes there is no need in that bit, it's semantic becomes subset of the IPOIB_FLAG_OPER_UP bit. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-04-02 18:39:04 +08:00
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return;
IPoIB: Start multicast join process only on active ports The driver starts the mcast_join task whenever the netdev interface is UP without relation to the underlying IB port state. Until the port state is ACTIVE all the join requests are irrelevant, and the IB core returns -EINVAL. So the user will see errors such as: "multicast join failed for ff12:401b:... , status -22". Instead, have ipoib_mcast_join_task() return when the port is not active. It will be called again when the port state is changed and the low-level driver triggers the IB_EVENT_PORT_ACTIVE event or the IB_EVENT_CLIENT_REREGISTER event. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:53 +08:00
if (ib_query_port(priv->ca, priv->port, &port_attr) ||
port_attr.state != IB_PORT_ACTIVE) {
ipoib_dbg(priv, "port state is not ACTIVE (state = %d) suspending join task\n",
port_attr.state);
return;
}
IPoIB: Remove unnecessary port query There are two queries for port attributes one after another. A second call is not needed since port_attr structure already holds the data. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Alex Estrin <alex.estrin@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-08-20 23:15:08 +08:00
priv->local_lid = port_attr.lid;
IPoIB: Start multicast join process only on active ports The driver starts the mcast_join task whenever the netdev interface is UP without relation to the underlying IB port state. Until the port state is ACTIVE all the join requests are irrelevant, and the IB core returns -EINVAL. So the user will see errors such as: "multicast join failed for ff12:401b:... , status -22". Instead, have ipoib_mcast_join_task() return when the port is not active. It will be called again when the port state is changed and the low-level driver triggers the IB_EVENT_PORT_ACTIVE event or the IB_EVENT_CLIENT_REREGISTER event. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:53 +08:00
IB/core: Add netdev and gid attributes paramteres to cache Adding an ability to query the IB cache by a netdev and get the attributes of a GID. These parameters are necessary in order to successfully resolve the required GID (when the netdevice is known) and get the Ethernet L2 attributes from a GID. Signed-off-by: Matan Barak <matanb@mellanox.com> Reviewed-By: Devesh Sharma <devesh.sharma@avagotech.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-10-15 23:38:45 +08:00
if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid, NULL))
IB/ipoib: Fix typos in error messages Trivial error message fixups. Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-05-18 21:12:54 +08:00
ipoib_warn(priv, "ib_query_gid() failed\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
else
memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid));
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
spin_lock_irq(&priv->lock);
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!priv->broadcast) {
IPoIB: Yet another fix for send-only joins Even after the last fix, it's still possible for a send-only join to start before the join for the broadcast group has finished. This could cause us to create a multicast group using attributes from the broadcast group that haven't been initialized yet, so we would use garbage for the Q_Key, etc. Fix this by waiting until the broadcast group's attached flag is set before starting send-only joins. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-12 04:22:12 +08:00
struct ipoib_mcast *broadcast;
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
broadcast = ipoib_mcast_alloc(dev, 0);
IPoIB: Yet another fix for send-only joins Even after the last fix, it's still possible for a send-only join to start before the join for the broadcast group has finished. This could cause us to create a multicast group using attributes from the broadcast group that haven't been initialized yet, so we would use garbage for the Q_Key, etc. Fix this by waiting until the broadcast group's attached flag is set before starting send-only joins. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-12 04:22:12 +08:00
if (!broadcast) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ipoib_warn(priv, "failed to allocate broadcast group\n");
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
* Restart us after a 1 second delay to retry
* creating our broadcast group and attaching to
* it. Until this succeeds, this ipoib dev is
* completely stalled (multicast wise).
*/
__ipoib_mcast_schedule_join_thread(priv, NULL, 1);
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IPoIB: Yet another fix for send-only joins Even after the last fix, it's still possible for a send-only join to start before the join for the broadcast group has finished. This could cause us to create a multicast group using attributes from the broadcast group that haven't been initialized yet, so we would use garbage for the Q_Key, etc. Fix this by waiting until the broadcast group's attached flag is set before starting send-only joins. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-12 04:22:12 +08:00
memcpy(broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sizeof (union ib_gid));
IPoIB: Yet another fix for send-only joins Even after the last fix, it's still possible for a send-only join to start before the join for the broadcast group has finished. This could cause us to create a multicast group using attributes from the broadcast group that haven't been initialized yet, so we would use garbage for the Q_Key, etc. Fix this by waiting until the broadcast group's attached flag is set before starting send-only joins. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-12 04:22:12 +08:00
priv->broadcast = broadcast;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
__ipoib_mcast_add(dev, priv->broadcast);
}
if (!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (IS_ERR_OR_NULL(priv->broadcast->mc) &&
!test_bit(IPOIB_MCAST_FLAG_BUSY, &priv->broadcast->flags)) {
mcast = priv->broadcast;
if (mcast->backoff > 1 &&
time_before(jiffies, mcast->delay_until)) {
delay_until = mcast->delay_until;
mcast = NULL;
}
}
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
* We'll never get here until the broadcast group is both allocated
* and attached
*/
list_for_each_entry(mcast, &priv->multicast_list, list) {
if (IS_ERR_OR_NULL(mcast->mc) &&
!test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags) &&
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
(!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) ||
!skb_queue_empty(&mcast->pkt_queue))) {
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (mcast->backoff == 1 ||
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
time_after_eq(jiffies, mcast->delay_until)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Found the next unjoined group */
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
init_completion(&mcast->done);
set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
spin_unlock_irq(&priv->lock);
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
ipoib_mcast_join(dev, mcast);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
spin_lock_irq(&priv->lock);
} else if (!delay_until ||
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
time_before(mcast->delay_until, delay_until))
delay_until = mcast->delay_until;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
}
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
mcast = NULL;
ipoib_dbg_mcast(priv, "successfully started all multicast joins\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
out:
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (delay_until) {
cancel_delayed_work(&priv->mcast_task);
queue_delayed_work(priv->wq, &priv->mcast_task,
delay_until - jiffies);
}
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (mcast) {
init_completion(&mcast->done);
set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
}
spin_unlock_irq(&priv->lock);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (mcast)
IB/ipoib: Clean up send-only multicast joins Even though we don't expect the group to be created by the SM we sill need to provide all the parameters to force the SM to validate they are correct. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-09-04 05:05:58 +08:00
ipoib_mcast_join(dev, mcast);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
int ipoib_mcast_start_thread(struct net_device *dev)
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
unsigned long flags;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ipoib_dbg_mcast(priv, "starting multicast thread\n");
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_lock_irqsave(&priv->lock, flags);
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
__ipoib_mcast_schedule_join_thread(priv, NULL, 0);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_unlock_irqrestore(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
IB/ipoib: No longer use flush as a parameter Various places in the IPoIB code had a deadlock related to flushing the ipoib workqueue. Now that we have per device workqueues and a specific flush workqueue, there is no longer a deadlock issue with flushing the device specific workqueues and we can do so unilaterally. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:04 +08:00
int ipoib_mcast_stop_thread(struct net_device *dev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
unsigned long flags;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ipoib_dbg_mcast(priv, "stopping multicast thread\n");
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_lock_irqsave(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
cancel_delayed_work(&priv->mcast_task);
IB/ipoib: drop mcast_mutex usage We needed the mcast_mutex when we had to prevent the join completion callback from having the value it stored in mcast->mc overwritten by a delayed return from ib_sa_join_multicast. By storing the return of ib_sa_join_multicast in an intermediate variable, we prevent a delayed return from ib_sa_join_multicast overwriting the valid contents of mcast->mc, and we no longer need a mutex to force the join callback to run after the return of ib_sa_join_multicast. This allows us to do away with the mutex entirely and protect our critical sections with a just a spinlock instead. This is highly desirable as there were some places where we couldn't use a mutex because the code was not allowed to sleep, and so we were currently using a mix of mutex and spinlock to protect what we needed to protect. Now we only have a spin lock and the locking complexity is greatly reduced. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:07 +08:00
spin_unlock_irqrestore(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: No longer use flush as a parameter Various places in the IPoIB code had a deadlock related to flushing the ipoib workqueue. Now that we have per device workqueues and a specific flush workqueue, there is no longer a deadlock issue with flushing the device specific workqueues and we can do so unilaterally. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:04 +08:00
flush_workqueue(priv->wq);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
IB/IPoIB: factor out common multicast list removal code Code cleanup to remove multicast specific code from ipoib_main.c The removal of a list of multicast groups occurs in three places. Create a new function ipoib_mcast_remove_list(). Use this new function in ipoib_main.c too. That in turn allows the dropping of two functions that were exported from ipoib_multicast.c for expiration of mc groups. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-12-21 22:42:53 +08:00
static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
int ret = 0;
IPoIB: Fix race in detaching from mcast group before attaching There's a race between ipoib_mcast_leave() and ipoib_mcast_join_finish() where we can try to detach from a multicast group before we've attached to it. Fix this by reordering the code in ipoib_mcast_leave to free the multicast group first, which waits for the multicast callback thread (which calls ipoib_mcast_join_finish()) to complete before detaching from the group. Signed-off-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-03-20 06:31:36 +08:00
if (test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
ipoib_warn(priv, "ipoib_mcast_leave on an in-flight join\n");
if (!IS_ERR_OR_NULL(mcast->mc))
IPoIB: Fix race in detaching from mcast group before attaching There's a race between ipoib_mcast_leave() and ipoib_mcast_join_finish() where we can try to detach from a multicast group before we've attached to it. Fix this by reordering the code in ipoib_mcast_leave to free the multicast group first, which waits for the multicast callback thread (which calls ipoib_mcast_join_finish()) to complete before detaching from the group. Signed-off-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-03-20 06:31:36 +08:00
ib_sa_free_multicast(mcast->mc);
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
if (test_and_clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "leaving MGID %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
/* Remove ourselves from the multicast group */
IPoIB: Get rid of ipoib_mcast_detach() wrapper ipoib_mcast_detach() does nothing except call ib_detach_mcast(), so just use the core API in the one place that does a multicast group detach. add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-105 (-105) function old new delta ipoib_mcast_leave 357 319 -38 ipoib_mcast_detach 67 - -67 Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-07-15 14:48:50 +08:00
ret = ib_detach_mcast(priv->qp, &mcast->mcmember.mgid,
be16_to_cpu(mcast->mcmember.mlid));
IB/sa: Track multicast join/leave requests The IB SA tracks multicast join/leave requests on a per port basis and does not do any reference counting: if two users of the same port join the same group, and one leaves that group, then the SA will remove the port from the group even though there is one user who wants to stay a member left. Therefore, in order to support multiple users of the same multicast group from the same port, we need to perform reference counting locally. To do this, add an multicast submodule to ib_sa to perform reference counting of multicast join/leave operations. Modify ib_ipoib (the only in-kernel user of multicast) to use the new interface. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-02-16 09:00:17 +08:00
if (ret)
IPoIB: Get rid of ipoib_mcast_detach() wrapper ipoib_mcast_detach() does nothing except call ib_detach_mcast(), so just use the core API in the one place that does a multicast group detach. add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-105 (-105) function old new delta ipoib_mcast_leave 357 319 -38 ipoib_mcast_detach 67 - -67 Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-07-15 14:48:50 +08:00
ipoib_warn(priv, "ib_detach_mcast failed (result = %d)\n", ret);
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
} else if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
ipoib_dbg(priv, "leaving with no mcmember but not a "
"SENDONLY join\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
IB/IPoIB: factor out common multicast list removal code Code cleanup to remove multicast specific code from ipoib_main.c The removal of a list of multicast groups occurs in three places. Create a new function ipoib_mcast_remove_list(). Use this new function in ipoib_main.c too. That in turn allows the dropping of two functions that were exported from ipoib_multicast.c for expiration of mc groups. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-12-21 22:42:53 +08:00
void ipoib_mcast_remove_list(struct net_device *dev, struct list_head *remove_list)
{
struct ipoib_mcast *mcast, *tmcast;
list_for_each_entry_safe(mcast, tmcast, remove_list, list) {
ipoib_mcast_leave(dev, mcast);
ipoib_mcast_free(mcast);
}
}
IPoIB: Use a private hash table for path lookup in xmit path Dave Miller <davem@davemloft.net> provided a detailed description of why the way IPoIB is using neighbours for its own ipoib_neigh struct is buggy: Any time an ipoib_neigh is changed, a sequence like the following is made: spin_lock_irqsave(&priv->lock, flags); /* * It's safe to call ipoib_put_ah() inside * priv->lock here, because we know that * path->ah will always hold one more reference, * so ipoib_put_ah() will never do more than * decrement the ref count. */ if (neigh->ah) ipoib_put_ah(neigh->ah); list_del(&neigh->list); ipoib_neigh_free(dev, neigh); spin_unlock_irqrestore(&priv->lock, flags); ipoib_path_lookup(skb, n, dev); This doesn't work, because you're leaving a stale pointer to the freed up ipoib_neigh in the special neigh->ha pointer cookie. Yes, it even fails with all the locking done to protect _changes_ to *ipoib_neigh(n), and with the code in ipoib_neigh_free() that NULLs out the pointer. The core issue is that read side calls to *to_ipoib_neigh(n) are not being synchronized at all, they are performed without any locking. So whether we hold the lock or not when making changes to *ipoib_neigh(n) you still can have threads see references to freed up ipoib_neigh objects. cpu 1 cpu 2 n = *ipoib_neigh() *ipoib_neigh() = NULL kfree(n) n->foo == OOPS [..] Perhaps the ipoib code can have a private path database it manages entirely itself, which holds all the necessary information and is looked up by some generic key which is available easily at transmit time and does not involve generic neighbour entries. See <http://marc.info/?l=linux-rdma&m=132812793105624&w=2> and <http://marc.info/?l=linux-rdma&w=2&r=1&s=allows+references+to+freed+memory&q=b> for the full discussion. This patch aims to solve the race conditions found in the IPoIB driver. The patch removes the connection between the core networking neighbour structure and the ipoib_neigh structure. In addition to avoiding the race described above, it allows us to handle SKBs carrying IP packets that don't have any associated neighbour. We add an ipoib_neigh hash table with N buckets where the key is the destination hardware address. The ipoib_neigh is fetched from the hash table and instead of the stashed location in the neighbour structure. The hash table uses both RCU and reference counting to guarantee that no ipoib_neigh instance is ever deleted while in use. Fetching the ipoib_neigh structure instance from the hash also makes the special code in ipoib_start_xmit that handles remote and local bonding failover redundant. Aged ipoib_neigh instances are deleted by a garbage collection task that runs every M seconds and deletes every ipoib_neigh instance that was idle for at least 2*M seconds. The deletion is safe since the ipoib_neigh instances are protected using RCU and reference count mechanisms. The number of buckets (N) and frequency of running the GC thread (M), are taken from the exported arb_tbl. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-25 01:05:22 +08:00
void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct ipoib_mcast *mcast;
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
unsigned long flags;
IPoIB: Use a private hash table for path lookup in xmit path Dave Miller <davem@davemloft.net> provided a detailed description of why the way IPoIB is using neighbours for its own ipoib_neigh struct is buggy: Any time an ipoib_neigh is changed, a sequence like the following is made: spin_lock_irqsave(&priv->lock, flags); /* * It's safe to call ipoib_put_ah() inside * priv->lock here, because we know that * path->ah will always hold one more reference, * so ipoib_put_ah() will never do more than * decrement the ref count. */ if (neigh->ah) ipoib_put_ah(neigh->ah); list_del(&neigh->list); ipoib_neigh_free(dev, neigh); spin_unlock_irqrestore(&priv->lock, flags); ipoib_path_lookup(skb, n, dev); This doesn't work, because you're leaving a stale pointer to the freed up ipoib_neigh in the special neigh->ha pointer cookie. Yes, it even fails with all the locking done to protect _changes_ to *ipoib_neigh(n), and with the code in ipoib_neigh_free() that NULLs out the pointer. The core issue is that read side calls to *to_ipoib_neigh(n) are not being synchronized at all, they are performed without any locking. So whether we hold the lock or not when making changes to *ipoib_neigh(n) you still can have threads see references to freed up ipoib_neigh objects. cpu 1 cpu 2 n = *ipoib_neigh() *ipoib_neigh() = NULL kfree(n) n->foo == OOPS [..] Perhaps the ipoib code can have a private path database it manages entirely itself, which holds all the necessary information and is looked up by some generic key which is available easily at transmit time and does not involve generic neighbour entries. See <http://marc.info/?l=linux-rdma&m=132812793105624&w=2> and <http://marc.info/?l=linux-rdma&w=2&r=1&s=allows+references+to+freed+memory&q=b> for the full discussion. This patch aims to solve the race conditions found in the IPoIB driver. The patch removes the connection between the core networking neighbour structure and the ipoib_neigh structure. In addition to avoiding the race described above, it allows us to handle SKBs carrying IP packets that don't have any associated neighbour. We add an ipoib_neigh hash table with N buckets where the key is the destination hardware address. The ipoib_neigh is fetched from the hash table and instead of the stashed location in the neighbour structure. The hash table uses both RCU and reference counting to guarantee that no ipoib_neigh instance is ever deleted while in use. Fetching the ipoib_neigh structure instance from the hash also makes the special code in ipoib_start_xmit that handles remote and local bonding failover redundant. Aged ipoib_neigh instances are deleted by a garbage collection task that runs every M seconds and deletes every ipoib_neigh instance that was idle for at least 2*M seconds. The deletion is safe since the ipoib_neigh instances are protected using RCU and reference count mechanisms. The number of buckets (N) and frequency of running the GC thread (M), are taken from the exported arb_tbl. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-25 01:05:22 +08:00
void *mgid = daddr + 4;
ipoib: Need to do dst_neigh_lookup_skb() outside of priv->lock. Otherwise local_bh_enable() complains. Reported-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-06 12:08:05 +08:00
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
spin_lock_irqsave(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IPoIB: Don't drop multicast sends when they can be queued When set_multicast_list() is called the multicast task is restarted and the IPOIB_MCAST_STARTED bit is cleared. As a result for some window of time, multicast packets are not transmitted nor queued but rather dropped by ipoib_mcast_send(). These dropped packets are painful in two cases: - bonding fail-over which both calls set_multicast_list() on the new active slave and sends Gratuitous ARP through that slave. - IP_DROP_MEMBERSHIP code which both calls set_multicast_list() on the device and issues IGMP leave. In both these cases, depending on the scheduling of the IPoIB multicast task, the packets would be dropped. As a result, in the bonding case, the failover would not be detected by the peers until their neighbour is renewed the neighbour (which takes a few tens of seconds). In the IGMP case, the IP router doesn't get an IGMP leave and would only learn on that from further probes on the group (also a delay of at least a few tens of seconds). Fix this by allowing transmission (or queuing) depending on the IPOIB_FLAG_OPER_UP flag instead of the IPOIB_MCAST_STARTED flag. Signed-off-by: Olga Shern <olgas@voltaire.com> Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-03-11 22:10:02 +08:00
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags) ||
IPoIB: Yet another fix for send-only joins Even after the last fix, it's still possible for a send-only join to start before the join for the broadcast group has finished. This could cause us to create a multicast group using attributes from the broadcast group that haven't been initialized yet, so we would use garbage for the Q_Key, etc. Fix this by waiting until the broadcast group's attached flag is set before starting send-only joins. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-12 04:22:12 +08:00
!priv->broadcast ||
!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
[IPoIB]: Convert to netdevice internal stats Use the stats member of struct netdevice in IPoIB, so we can save memory by deleting the stats member of struct ipoib_dev_priv, and save code by deleting ipoib_get_stats(). Signed-off-by: Roland Dreier <rolandd@cisco.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-09-29 06:33:51 +08:00
++dev->stats.tx_dropped;
IPoIB: Don't start send-only joins while multicast thread is stopped Fix the following race scenario: - Device is up. - Port event or set mcast list triggers ipoib_mcast_stop_thread, this cancels the query and waits on mcast "done" completion. - Completion is called and "done" is set. - Meanwhile, ipoib_mcast_send arrives and starts a new query, re-initializing "done". Fix this by adding a "multicast started" bit and checking it before starting a send-only join. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-08 08:37:08 +08:00
dev_kfree_skb_any(skb);
goto unlock;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
mcast = __ipoib_mcast_find(dev, mgid);
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (!mcast || !mcast->ah) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!mcast) {
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
/* Let's create a new send only group now */
ipoib_dbg_mcast(priv, "setting up send only multicast group for %pI6\n",
mgid);
mcast = ipoib_mcast_alloc(dev, 0);
if (!mcast) {
ipoib_warn(priv, "unable to allocate memory "
"for multicast structure\n");
++dev->stats.tx_dropped;
dev_kfree_skb_any(skb);
goto unlock;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
set_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags);
memcpy(mcast->mcmember.mgid.raw, mgid,
sizeof (union ib_gid));
__ipoib_mcast_add(dev, mcast);
list_add_tail(&mcast->list, &priv->multicast_list);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE)
skb_queue_tail(&mcast->pkt_queue, skb);
IPoIB: Lock accesses to multicast packet queues Avoid corrupting mcast->pkt_queue by serializing access with priv->tx_lock. Also, update dropped packet statistics to count multicast packets removed from pkt_queue as dropped. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-18 04:19:40 +08:00
else {
[IPoIB]: Convert to netdevice internal stats Use the stats member of struct netdevice in IPoIB, so we can save memory by deleting the stats member of struct ipoib_dev_priv, and save code by deleting ipoib_get_stats(). Signed-off-by: Roland Dreier <rolandd@cisco.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-09-29 06:33:51 +08:00
++dev->stats.tx_dropped;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
dev_kfree_skb_any(skb);
IPoIB: Lock accesses to multicast packet queues Avoid corrupting mcast->pkt_queue by serializing access with priv->tx_lock. Also, update dropped packet statistics to count multicast packets removed from pkt_queue as dropped. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-18 04:19:40 +08:00
}
IB/ipoib: deserialize multicast joins Allow the ipoib layer to attempt to join all outstanding multicast groups at once. The ib_sa layer will serialize multiple attempts to join the same group, but will process attempts to join different groups in parallel. Take advantage of that. In order to make this happen, change the mcast_join_thread to loop through all needed joins, sending a join request for each one that we still need to join. There are a few special cases we handle though: 1) Don't attempt to join anything but the broadcast group until the join of the broadcast group has succeeded. 2) No longer restart the join task at the end of completion handling. If we completed successfully, we are done. The join task now needs kicked either by mcast_send or mcast_restart_task or mcast_start_thread, but should not need started anytime else except when scheduling a backoff attempt to rejoin. 3) No longer use separate join/completion routines for regular and sendonly joins, pass them all through the same routine and just do the right thing based on the SENDONLY join flag. 4) Only try to join a SENDONLY join twice, then drop the packets and quit trying. We leave the mcast group in the list so that if we get a new packet, all that we have to do is queue up the packet and restart the join task and it will automatically try to join twice and then either send or flush the queue again. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:06 +08:00
if (!test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)) {
__ipoib_mcast_schedule_join_thread(priv, NULL, 0);
}
} else {
IPoIB: Use a private hash table for path lookup in xmit path Dave Miller <davem@davemloft.net> provided a detailed description of why the way IPoIB is using neighbours for its own ipoib_neigh struct is buggy: Any time an ipoib_neigh is changed, a sequence like the following is made: spin_lock_irqsave(&priv->lock, flags); /* * It's safe to call ipoib_put_ah() inside * priv->lock here, because we know that * path->ah will always hold one more reference, * so ipoib_put_ah() will never do more than * decrement the ref count. */ if (neigh->ah) ipoib_put_ah(neigh->ah); list_del(&neigh->list); ipoib_neigh_free(dev, neigh); spin_unlock_irqrestore(&priv->lock, flags); ipoib_path_lookup(skb, n, dev); This doesn't work, because you're leaving a stale pointer to the freed up ipoib_neigh in the special neigh->ha pointer cookie. Yes, it even fails with all the locking done to protect _changes_ to *ipoib_neigh(n), and with the code in ipoib_neigh_free() that NULLs out the pointer. The core issue is that read side calls to *to_ipoib_neigh(n) are not being synchronized at all, they are performed without any locking. So whether we hold the lock or not when making changes to *ipoib_neigh(n) you still can have threads see references to freed up ipoib_neigh objects. cpu 1 cpu 2 n = *ipoib_neigh() *ipoib_neigh() = NULL kfree(n) n->foo == OOPS [..] Perhaps the ipoib code can have a private path database it manages entirely itself, which holds all the necessary information and is looked up by some generic key which is available easily at transmit time and does not involve generic neighbour entries. See <http://marc.info/?l=linux-rdma&m=132812793105624&w=2> and <http://marc.info/?l=linux-rdma&w=2&r=1&s=allows+references+to+freed+memory&q=b> for the full discussion. This patch aims to solve the race conditions found in the IPoIB driver. The patch removes the connection between the core networking neighbour structure and the ipoib_neigh structure. In addition to avoiding the race described above, it allows us to handle SKBs carrying IP packets that don't have any associated neighbour. We add an ipoib_neigh hash table with N buckets where the key is the destination hardware address. The ipoib_neigh is fetched from the hash table and instead of the stashed location in the neighbour structure. The hash table uses both RCU and reference counting to guarantee that no ipoib_neigh instance is ever deleted while in use. Fetching the ipoib_neigh structure instance from the hash also makes the special code in ipoib_start_xmit that handles remote and local bonding failover redundant. Aged ipoib_neigh instances are deleted by a garbage collection task that runs every M seconds and deletes every ipoib_neigh instance that was idle for at least 2*M seconds. The deletion is safe since the ipoib_neigh instances are protected using RCU and reference count mechanisms. The number of buckets (N) and frequency of running the GC thread (M), are taken from the exported arb_tbl. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-25 01:05:22 +08:00
struct ipoib_neigh *neigh;
spin_unlock_irqrestore(&priv->lock, flags);
neigh = ipoib_neigh_get(dev, daddr);
spin_lock_irqsave(&priv->lock, flags);
if (!neigh) {
neigh = ipoib_neigh_alloc(daddr, dev);
if (neigh) {
kref_get(&mcast->ah->ref);
neigh->ah = mcast->ah;
list_add_tail(&neigh->list, &mcast->neigh_list);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
}
IPoIB: Drop priv->lock before calling ipoib_send() IPoIB currently must use irqsave locking for priv->lock, since it is taken from interrupt context in one path. However, ipoib_send() does skb_orphan(), and the network stack locking is not IRQ-safe. Therefore we need to make sure we don't hold priv->lock when calling ipoib_send() to avoid lockdep warnings (the code was almost certainly safe in practice, since the only code path that takes priv->lock from interrupt context would never call into the network stack). Addresses: http://bugzilla.kernel.org/show_bug.cgi?id=13757 Reported-by: Bart Van Assche <bart.vanassche@gmail.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-06 11:23:40 +08:00
spin_unlock_irqrestore(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ipoib_send(dev, skb, mcast->ah, IB_MULTICAST_QPN);
IPoIB: Use a private hash table for path lookup in xmit path Dave Miller <davem@davemloft.net> provided a detailed description of why the way IPoIB is using neighbours for its own ipoib_neigh struct is buggy: Any time an ipoib_neigh is changed, a sequence like the following is made: spin_lock_irqsave(&priv->lock, flags); /* * It's safe to call ipoib_put_ah() inside * priv->lock here, because we know that * path->ah will always hold one more reference, * so ipoib_put_ah() will never do more than * decrement the ref count. */ if (neigh->ah) ipoib_put_ah(neigh->ah); list_del(&neigh->list); ipoib_neigh_free(dev, neigh); spin_unlock_irqrestore(&priv->lock, flags); ipoib_path_lookup(skb, n, dev); This doesn't work, because you're leaving a stale pointer to the freed up ipoib_neigh in the special neigh->ha pointer cookie. Yes, it even fails with all the locking done to protect _changes_ to *ipoib_neigh(n), and with the code in ipoib_neigh_free() that NULLs out the pointer. The core issue is that read side calls to *to_ipoib_neigh(n) are not being synchronized at all, they are performed without any locking. So whether we hold the lock or not when making changes to *ipoib_neigh(n) you still can have threads see references to freed up ipoib_neigh objects. cpu 1 cpu 2 n = *ipoib_neigh() *ipoib_neigh() = NULL kfree(n) n->foo == OOPS [..] Perhaps the ipoib code can have a private path database it manages entirely itself, which holds all the necessary information and is looked up by some generic key which is available easily at transmit time and does not involve generic neighbour entries. See <http://marc.info/?l=linux-rdma&m=132812793105624&w=2> and <http://marc.info/?l=linux-rdma&w=2&r=1&s=allows+references+to+freed+memory&q=b> for the full discussion. This patch aims to solve the race conditions found in the IPoIB driver. The patch removes the connection between the core networking neighbour structure and the ipoib_neigh structure. In addition to avoiding the race described above, it allows us to handle SKBs carrying IP packets that don't have any associated neighbour. We add an ipoib_neigh hash table with N buckets where the key is the destination hardware address. The ipoib_neigh is fetched from the hash table and instead of the stashed location in the neighbour structure. The hash table uses both RCU and reference counting to guarantee that no ipoib_neigh instance is ever deleted while in use. Fetching the ipoib_neigh structure instance from the hash also makes the special code in ipoib_start_xmit that handles remote and local bonding failover redundant. Aged ipoib_neigh instances are deleted by a garbage collection task that runs every M seconds and deletes every ipoib_neigh instance that was idle for at least 2*M seconds. The deletion is safe since the ipoib_neigh instances are protected using RCU and reference count mechanisms. The number of buckets (N) and frequency of running the GC thread (M), are taken from the exported arb_tbl. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-07-25 01:05:22 +08:00
if (neigh)
ipoib_neigh_put(neigh);
IPoIB: Drop priv->lock before calling ipoib_send() IPoIB currently must use irqsave locking for priv->lock, since it is taken from interrupt context in one path. However, ipoib_send() does skb_orphan(), and the network stack locking is not IRQ-safe. Therefore we need to make sure we don't hold priv->lock when calling ipoib_send() to avoid lockdep warnings (the code was almost certainly safe in practice, since the only code path that takes priv->lock from interrupt context would never call into the network stack). Addresses: http://bugzilla.kernel.org/show_bug.cgi?id=13757 Reported-by: Bart Van Assche <bart.vanassche@gmail.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-06 11:23:40 +08:00
return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
IPoIB: Don't start send-only joins while multicast thread is stopped Fix the following race scenario: - Device is up. - Port event or set mcast list triggers ipoib_mcast_stop_thread, this cancels the query and waits on mcast "done" completion. - Completion is called and "done" is set. - Meanwhile, ipoib_mcast_send arrives and starts a new query, re-initializing "done". Fix this by adding a "multicast started" bit and checking it before starting a send-only join. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-02-08 08:37:08 +08:00
unlock:
IPoIB: Use netif_tx_lock() and get rid of private tx_lock, LLTX Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling tx_lock. Not only do we want to get rid of LLTX, this actually causes problems because of the skb_orphan() done with this tx_lock held: some skb destructors expect to be run with interrupts enabled. The simplest fix for this is to get rid of the driver-private tx_lock and stop using LLTX. We kill off priv->tx_lock and use netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit tricky because we need to update places that take priv->lock inside the tx_lock to disable IRQs, rather than relying on tx_lock having already disabled IRQs. Also, there are a couple of places where we need to disable BHs to make sure we have a consistent context to call netif_tx_lock() (since we no longer can use _irqsave() variants), and we also have to change ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather than directly, because ipoib_send_comp_handler() runs in interrupt context and drain_tx_cq() must run in BH context so it can call netif_tx_lock(). Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-10-01 01:36:21 +08:00
spin_unlock_irqrestore(&priv->lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
void ipoib_mcast_dev_flush(struct net_device *dev)
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
LIST_HEAD(remove_list);
IPoIB: Fix memory leak of multicast group structures The current handling of multicast groups in IPoIB ends up never freeing send-only multicast groups. It turns out the logic was much more complicated than it needed to be; we can fix this bug and completely kill ipoib_mcast_dev_down() at the same time. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-13 06:32:20 +08:00
struct ipoib_mcast *mcast, *tmcast;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
unsigned long flags;
ipoib_dbg_mcast(priv, "flushing multicast list\n");
spin_lock_irqsave(&priv->lock, flags);
IPoIB: Fix memory leak of multicast group structures The current handling of multicast groups in IPoIB ends up never freeing send-only multicast groups. It turns out the logic was much more complicated than it needed to be; we can fix this bug and completely kill ipoib_mcast_dev_down() at the same time. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-13 06:32:20 +08:00
list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
list_del(&mcast->list);
rb_erase(&mcast->rb_node, &priv->multicast_tree);
list_add_tail(&mcast->list, &remove_list);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
if (priv->broadcast) {
IB: Whitespace fixes Remove some trailing whitespace that has snuck in despite the best efforts of whitespace=error-all. Also fix a few other whitespace bogosities. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-09-23 06:22:46 +08:00
rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree);
IPoIB: Fix memory leak of multicast group structures The current handling of multicast groups in IPoIB ends up never freeing send-only multicast groups. It turns out the logic was much more complicated than it needed to be; we can fix this bug and completely kill ipoib_mcast_dev_down() at the same time. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-13 06:32:20 +08:00
list_add_tail(&priv->broadcast->list, &remove_list);
priv->broadcast = NULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
spin_unlock_irqrestore(&priv->lock, flags);
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
* make sure the in-flight joins have finished before we attempt
* to leave
*/
IPoIB: Fix usage of uninitialized multicast objects The driver should avoid calling ib_sa_free_multicast on the mcast->mc object until it finishes its initialization state. Otherwise we can crash when ipoib_mcast_dev_flush() attempts to use the uninitialized multicast object. Instead, only call wait_for_completion() for multicast entries that started the join process, meaning that ib_sa_join_multicast() finished. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:51 +08:00
list_for_each_entry_safe(mcast, tmcast, &remove_list, list)
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
IPoIB: Fix usage of uninitialized multicast objects The driver should avoid calling ib_sa_free_multicast on the mcast->mc object until it finishes its initialization state. Otherwise we can crash when ipoib_mcast_dev_flush() attempts to use the uninitialized multicast object. Instead, only call wait_for_completion() for multicast entries that started the join process, meaning that ib_sa_join_multicast() finished. Signed-off-by: Erez Shitrit <erezsh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2013-10-16 22:37:51 +08:00
wait_for_completion(&mcast->done);
IB/IPoIB: factor out common multicast list removal code Code cleanup to remove multicast specific code from ipoib_main.c The removal of a list of multicast groups occurs in three places. Create a new function ipoib_mcast_remove_list(). Use this new function in ipoib_main.c too. That in turn allows the dropping of two functions that were exported from ipoib_multicast.c for expiration of mc groups. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-12-21 22:42:53 +08:00
ipoib_mcast_remove_list(dev, &remove_list);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
ipoib: remove addrlen check for mc addresses Finally this bit can be removed. Currently, after the bonding driver is changed/fixed (32a806c194ea112cfab00f558482dd97bee5e44e net-next-2.6), that's not possible for an addr with different length than dev->addr_len to be present in list. Removing this check as in new mc_list there will be no addrlen in the record. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-03-22 11:21:39 +08:00
static int ipoib_mcast_addr_is_valid(const u8 *addr, const u8 *broadcast)
IPoIB: Check multicast address format Check that the format of multicast link addresses is correct before taking them from dev->mc_list to priv->multicast_list. This way we never try to send a bogus address to the SA, which prevents badness from erronous 'ip maddr addr add', broken bonding drivers, etc. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-06 11:23:40 +08:00
{
/* reserved QPN, prefix, scope */
if (memcmp(addr, broadcast, 6))
return 0;
/* signature lower, pkey */
if (memcmp(addr + 7, broadcast + 7, 3))
return 0;
return 1;
}
WorkStruct: make allyesconfig Fix up for make allyesconfig. Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 22:57:56 +08:00
void ipoib_mcast_restart_task(struct work_struct *work)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
WorkStruct: make allyesconfig Fix up for make allyesconfig. Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 22:57:56 +08:00
struct ipoib_dev_priv *priv =
container_of(work, struct ipoib_dev_priv, restart_task);
struct net_device *dev = priv->dev;
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-02 05:22:57 +08:00
struct netdev_hw_addr *ha;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct ipoib_mcast *mcast, *tmcast;
LIST_HEAD(remove_list);
unsigned long flags;
IPoIB: Allow setting policy to ignore multicast groups The kernel IB stack allows (through the RDMA CM) userspace applications to join and use multicast groups from the IPoIB MGID range. This allows multicast traffic to be handled directly from userspace QPs, without going through the kernel stack, which gives better performance for some applications. However, to fully interoperate with IP multicast, such userspace applications need to participate in IGMP reports and queries, or else routers may not forward the multicast traffic to the system where the application is running. The simplest way to do this is to share the kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to join multicast groups that are being handled directly in userspace. However, in such cases, the actual multicast traffic should not also be handled by the IPoIB interface, because that would burn resources handling multicast packets that will just be discarded in the kernel. To handle this, this patch adds lookup on the database used for IB multicast group reference counting when IPoIB is joining multicast groups, and if a multicast group is already handled by user space, then the IPoIB kernel driver ignores the group. This is controlled by a per-interface policy flag. When the flag is set, IPoIB will not join and attach its QP to a multicast group which already has an entry in the database; when the flag is cleared, IPoIB will behave as before this change. For each IPoIB interface, the /sys/class/net/$intf/umcast attribute controls the policy flag. The default value is off/0. Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 16:13:00 +08:00
struct ib_sa_mcmember_rec rec;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
/*
* shortcut...on shutdown flush is called next, just
* let it do all the work
*/
return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
ipoib_dbg_mcast(priv, "restarting multicast task\n");
Revert "IPoIB: Make ipoib_mcast_stop_thread flush the workqueue" This reverts commit bb42a6dd02fb2901a69dbec2358810735b14b186. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:38:46 +08:00
[NET]: Add netif_tx_lock Various drivers use xmit_lock internally to synchronise with their transmission routines. They do so without setting xmit_lock_owner. This is fine as long as netpoll is not in use. With netpoll it is possible for deadlocks to occur if xmit_lock_owner isn't set. This is because if a printk occurs while xmit_lock is held and xmit_lock_owner is not set can cause netpoll to attempt to take xmit_lock recursively. While it is possible to resolve this by getting netpoll to use trylock, it is suboptimal because netpoll's sole objective is to maximise the chance of getting the printk out on the wire. So delaying or dropping the message is to be avoided as much as possible. So the only alternative is to always set xmit_lock_owner. The following patch does this by introducing the netif_tx_lock family of functions that take care of setting/unsetting xmit_lock_owner. I renamed xmit_lock to _xmit_lock to indicate that it should not be used directly. I didn't provide irq versions of the netif_tx_lock functions since xmit_lock is meant to be a BH-disabling lock. This is pretty much a straight text substitution except for a small bug fix in winbond. It currently uses netif_stop_queue/spin_unlock_wait to stop transmission. This is unsafe as an IRQ can potentially wake up the queue. So it is safer to use netif_tx_disable. The hamradio bits used spin_lock_irq but it is unnecessary as xmit_lock must never be taken in an IRQ handler. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-10 03:20:56 +08:00
local_irq_save(flags);
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 15:13:44 +08:00
netif_addr_lock(dev);
IPoIB: Take dev->xmit_lock around mc_list accesses dev->mc_list accesses must be protected by dev->xmit_lock. Found by Eli Cohen <eli@mellanox.co.il>. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-12 03:47:34 +08:00
spin_lock(&priv->lock);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/*
* Unfortunately, the networking core only gives us a list of all of
* the multicast hardware addresses. We need to figure out which ones
* are new and which ones have been removed
*/
/* Clear out the found flag */
list_for_each_entry(mcast, &priv->multicast_list, list)
clear_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags);
/* Mark all of the entries that are found or don't exist */
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-02 05:22:57 +08:00
netdev_for_each_mc_addr(ha, dev) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
union ib_gid mgid;
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-02 05:22:57 +08:00
if (!ipoib_mcast_addr_is_valid(ha->addr, dev->broadcast))
IPoIB: Check multicast address format Check that the format of multicast link addresses is correct before taking them from dev->mc_list to priv->multicast_list. This way we never try to send a bogus address to the SA, which prevents badness from erronous 'ip maddr addr add', broken bonding drivers, etc. Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-09-06 11:23:40 +08:00
continue;
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-02 05:22:57 +08:00
memcpy(mgid.raw, ha->addr + 4, sizeof mgid);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
mcast = __ipoib_mcast_find(dev, &mgid);
if (!mcast || test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
struct ipoib_mcast *nmcast;
IPoIB: Allow setting policy to ignore multicast groups The kernel IB stack allows (through the RDMA CM) userspace applications to join and use multicast groups from the IPoIB MGID range. This allows multicast traffic to be handled directly from userspace QPs, without going through the kernel stack, which gives better performance for some applications. However, to fully interoperate with IP multicast, such userspace applications need to participate in IGMP reports and queries, or else routers may not forward the multicast traffic to the system where the application is running. The simplest way to do this is to share the kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to join multicast groups that are being handled directly in userspace. However, in such cases, the actual multicast traffic should not also be handled by the IPoIB interface, because that would burn resources handling multicast packets that will just be discarded in the kernel. To handle this, this patch adds lookup on the database used for IB multicast group reference counting when IPoIB is joining multicast groups, and if a multicast group is already handled by user space, then the IPoIB kernel driver ignores the group. This is controlled by a per-interface policy flag. When the flag is set, IPoIB will not join and attach its QP to a multicast group which already has an entry in the database; when the flag is cleared, IPoIB will behave as before this change. For each IPoIB interface, the /sys/class/net/$intf/umcast attribute controls the policy flag. The default value is off/0. Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 16:13:00 +08:00
/* ignore group which is directly joined by userspace */
if (test_bit(IPOIB_FLAG_UMCAST, &priv->flags) &&
!ib_sa_get_mcmember_rec(priv->ca, priv->port, &mgid, &rec)) {
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "ignoring multicast entry for mgid %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mgid.raw);
IPoIB: Allow setting policy to ignore multicast groups The kernel IB stack allows (through the RDMA CM) userspace applications to join and use multicast groups from the IPoIB MGID range. This allows multicast traffic to be handled directly from userspace QPs, without going through the kernel stack, which gives better performance for some applications. However, to fully interoperate with IP multicast, such userspace applications need to participate in IGMP reports and queries, or else routers may not forward the multicast traffic to the system where the application is running. The simplest way to do this is to share the kernel IGMP implementation by using the IP_ADD_MEMBERSHIP option to join multicast groups that are being handled directly in userspace. However, in such cases, the actual multicast traffic should not also be handled by the IPoIB interface, because that would burn resources handling multicast packets that will just be discarded in the kernel. To handle this, this patch adds lookup on the database used for IB multicast group reference counting when IPoIB is joining multicast groups, and if a multicast group is already handled by user space, then the IPoIB kernel driver ignores the group. This is controlled by a per-interface policy flag. When the flag is set, IPoIB will not join and attach its QP to a multicast group which already has an entry in the database; when the flag is cleared, IPoIB will behave as before this change. For each IPoIB interface, the /sys/class/net/$intf/umcast attribute controls the policy flag. The default value is off/0. Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2007-10-08 16:13:00 +08:00
continue;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Not found or send-only group, let's add a new entry */
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "adding multicast entry for mgid %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
nmcast = ipoib_mcast_alloc(dev, 0);
if (!nmcast) {
ipoib_warn(priv, "unable to allocate memory for multicast structure\n");
continue;
}
set_bit(IPOIB_MCAST_FLAG_FOUND, &nmcast->flags);
nmcast->mcmember.mgid = mgid;
if (mcast) {
/* Destroy the send only entry */
[PATCH] drivers: use list_move() This patch converts the combination of list_del(A) and list_add(A, B) to list_move(A, B) under drivers/. Acked-by: Corey Minyard <minyard@mvista.com> Cc: Ben Collins <bcollins@debian.org> Acked-by: Roland Dreier <rolandd@cisco.com> Cc: Alasdair Kergon <dm-devel@redhat.com> Cc: Gerd Knorr <kraxel@bytesex.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Frank Pavlic <fpavlic@de.ibm.com> Acked-by: Matthew Wilcox <matthew@wil.cx> Cc: Andrew Vasquez <linux-driver@qlogic.com> Cc: Mikael Starvik <starvik@axis.com> Cc: Greg Kroah-Hartman <greg@kroah.com> Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-26 15:24:41 +08:00
list_move_tail(&mcast->list, &remove_list);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rb_replace_node(&mcast->rb_node,
&nmcast->rb_node,
&priv->multicast_tree);
} else
__ipoib_mcast_add(dev, nmcast);
list_add_tail(&nmcast->list, &priv->multicast_list);
}
if (mcast)
set_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags);
}
/* Remove all of the entries don't exist anymore */
list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
if (!test_bit(IPOIB_MCAST_FLAG_FOUND, &mcast->flags) &&
!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
net: replace %p6 with %pI6 Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-30 03:52:50 +08:00
ipoib_dbg_mcast(priv, "deleting multicast group %pI6\n",
infiniband: ipoib replace IPOIB_GID_FMT with %p6 Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG() Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-29 13:37:22 +08:00
mcast->mcmember.mgid.raw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rb_erase(&mcast->rb_node, &priv->multicast_tree);
/* Move to the remove list */
[PATCH] drivers: use list_move() This patch converts the combination of list_del(A) and list_add(A, B) to list_move(A, B) under drivers/. Acked-by: Corey Minyard <minyard@mvista.com> Cc: Ben Collins <bcollins@debian.org> Acked-by: Roland Dreier <rolandd@cisco.com> Cc: Alasdair Kergon <dm-devel@redhat.com> Cc: Gerd Knorr <kraxel@bytesex.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Frank Pavlic <fpavlic@de.ibm.com> Acked-by: Matthew Wilcox <matthew@wil.cx> Cc: Andrew Vasquez <linux-driver@qlogic.com> Cc: Mikael Starvik <starvik@axis.com> Cc: Greg Kroah-Hartman <greg@kroah.com> Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-26 15:24:41 +08:00
list_move_tail(&mcast->list, &remove_list);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
}
IPoIB: Take dev->xmit_lock around mc_list accesses dev->mc_list accesses must be protected by dev->xmit_lock. Found by Eli Cohen <eli@mellanox.co.il>. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2006-01-12 03:47:34 +08:00
spin_unlock(&priv->lock);
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 15:13:44 +08:00
netif_addr_unlock(dev);
[NET]: Add netif_tx_lock Various drivers use xmit_lock internally to synchronise with their transmission routines. They do so without setting xmit_lock_owner. This is fine as long as netpoll is not in use. With netpoll it is possible for deadlocks to occur if xmit_lock_owner isn't set. This is because if a printk occurs while xmit_lock is held and xmit_lock_owner is not set can cause netpoll to attempt to take xmit_lock recursively. While it is possible to resolve this by getting netpoll to use trylock, it is suboptimal because netpoll's sole objective is to maximise the chance of getting the printk out on the wire. So delaying or dropping the message is to be avoided as much as possible. So the only alternative is to always set xmit_lock_owner. The following patch does this by introducing the netif_tx_lock family of functions that take care of setting/unsetting xmit_lock_owner. I renamed xmit_lock to _xmit_lock to indicate that it should not be used directly. I didn't provide irq versions of the netif_tx_lock functions since xmit_lock is meant to be a BH-disabling lock. This is pretty much a straight text substitution except for a small bug fix in winbond. It currently uses netif_stop_queue/spin_unlock_wait to stop transmission. This is unsafe as an IRQ can potentially wake up the queue. So it is safer to use netif_tx_disable. The hamradio bits used spin_lock_irq but it is unnecessary as xmit_lock must never be taken in an IRQ handler. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-10 03:20:56 +08:00
local_irq_restore(flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
* make sure the in-flight joins have finished before we attempt
* to leave
*/
list_for_each_entry_safe(mcast, tmcast, &remove_list, list)
if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
wait_for_completion(&mcast->done);
IB/IPoIB: factor out common multicast list removal code Code cleanup to remove multicast specific code from ipoib_main.c The removal of a list of multicast groups occurs in three places. Create a new function ipoib_mcast_remove_list(). Use this new function in ipoib_main.c too. That in turn allows the dropping of two functions that were exported from ipoib_multicast.c for expiration of mc groups. Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Christoph Lameter <cl@linux.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-12-21 22:42:53 +08:00
ipoib_mcast_remove_list(mcast->dev, &remove_list);
Revert "IPoIB: fix mcast_dev_flush/mcast_restart_task race" This reverts commit e5d1dcf1b0951f4ba00d93653942dda6196109d8. The series of IPoIB bug fixes that went into 3.19-rc1 introduce regressions, and after trying to sort things out, we decided to revert to 3.18's IPoIB driver and get things right for 3.20. Signed-off-by: Roland Dreier <roland@purestorage.com>
2015-01-31 07:39:11 +08:00
IB/ipoib: fix MCAST_FLAG_BUSY usage Commit a9c8ba5884 ("IPoIB: Fix usage of uninitialized multicast objects") added a new flag MCAST_JOIN_STARTED, but was not very strict in how it was used. We didn't always initialize the completion struct before we set the flag, and we didn't always call complete on the completion struct from all paths that complete it. And when we did complete it, sometimes we continued to touch the mcast entry after the completion, opening us up to possible use after free issues. This made it less than totally effective, and certainly made its use confusing. And in the flush function we would use the presence of this flag to signal that we should wait on the completion struct, but we never cleared this flag, ever. In order to make things clearer and aid in resolving the rtnl deadlock bug I've been chasing, I cleaned this up a bit. 1) Remove the MCAST_JOIN_STARTED flag entirely 2) Change MCAST_FLAG_BUSY so it now only means a join is in-flight 3) Test mcast->mc directly to see if we have completed ib_sa_join_multicast (using IS_ERR_OR_NULL) 4) Make sure that before setting MCAST_FLAG_BUSY we always initialize the mcast->done completion struct 5) Make sure that before calling complete(&mcast->done), we always clear the MCAST_FLAG_BUSY bit 6) Take the mcast_mutex before we call ib_sa_multicast_join and also take the mutex in our join callback. This forces ib_sa_multicast_join to return and set mcast->mc before we process the callback. This way, our callback can safely clear mcast->mc if there is an error on the join and we will do the right thing as a result in mcast_dev_flush. 7) Because we need the mutex to synchronize mcast->mc, we can no longer call mcast_sendonly_join directly from mcast_send and instead must add sendonly join processing to the mcast_join_task 8) Make MCAST_RUN mean that we have a working mcast subsystem, not that we have a running task. We know when we need to reschedule our join task thread and don't need a flag to tell us. 9) Add a helper for rescheduling the join task thread A number of different races are resolved with these changes. These races existed with the old MCAST_FLAG_BUSY usage, the MCAST_JOIN_STARTED flag was an attempt to address them, and while it helped, a determined effort could still trip things up. One race looks something like this: Thread 1 Thread 2 ib_sa_join_multicast (as part of running restart mcast task) alloc member call callback ifconfig ib0 down wait_for_completion callback call completes wait_for_completion in mcast_dev_flush completes mcast->mc is PTR_ERR_OR_NULL so we skip ib_sa_leave_multicast return from callback return from ib_sa_join_multicast set mcast->mc = return from ib_sa_multicast We now have a permanently unbalanced join/leave issue that trips up the refcounting in core/multicast.c Another like this: Thread 1 Thread 2 Thread 3 ib_sa_multicast_join ifconfig ib0 down priv->broadcast = NULL join_complete wait_for_completion mcast->mc is not yet set, so don't clear return from ib_sa_join_multicast and set mcast->mc complete return -EAGAIN (making mcast->mc invalid) call ib_sa_multicast_leave on invalid mcast->mc, hang forever By holding the mutex around ib_sa_multicast_join and taking the mutex early in the callback, we force mcast->mc to be valid at the time we run the callback. This allows us to clear mcast->mc if there is an error and the join is going to fail. We do this before we complete the mcast. In this way, mcast_dev_flush always sees consistent state in regards to mcast->mc membership at the time that the wait_for_completion() returns. Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-02-22 08:27:05 +08:00
/*
* Double check that we are still up
*/
if (test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) {
spin_lock_irqsave(&priv->lock, flags);
__ipoib_mcast_schedule_join_thread(priv, NULL, 0);
spin_unlock_irqrestore(&priv->lock, flags);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
[IPoIB] don't compile debug code if debugging isn't enabled Don't build ipoib_mcast_iter_ functions if CONFIG_INFINIBAND_IPOIB_DEBUG is not enabled -- their only callers will not be built either. Also move the prototype for ipoib_open() to ipoib.h to fix a sparse warning. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-03 12:51:01 +08:00
#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev)
{
struct ipoib_mcast_iter *iter;
iter = kmalloc(sizeof *iter, GFP_KERNEL);
if (!iter)
return NULL;
iter->dev = dev;
[IPoIB] add path record information in debugfs Add ibX_path files to debugfs that contain information about the IPoIB path cache. IPoIB ARP only gives GIDs, which the IPoIB driver must resolve to real IB paths through the ib_sa module. For debugging, when the ARP table looks OK but traffic isn't flowing, it's useful to be able to see if the resolution from GID to path worked. Also clean up the formatting of the existing _mcg debugfs files. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-08 02:33:11 +08:00
memset(iter->mgid.raw, 0, 16);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (ipoib_mcast_iter_next(iter)) {
[IPoIB] add path record information in debugfs Add ibX_path files to debugfs that contain information about the IPoIB path cache. IPoIB ARP only gives GIDs, which the IPoIB driver must resolve to real IB paths through the ib_sa module. For debugging, when the ARP table looks OK but traffic isn't flowing, it's useful to be able to see if the resolution from GID to path worked. Also clean up the formatting of the existing _mcg debugfs files. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-08 02:33:11 +08:00
kfree(iter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return NULL;
}
return iter;
}
int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter)
{
struct ipoib_dev_priv *priv = netdev_priv(iter->dev);
struct rb_node *n;
struct ipoib_mcast *mcast;
int ret = 1;
spin_lock_irq(&priv->lock);
n = rb_first(&priv->multicast_tree);
while (n) {
mcast = rb_entry(n, struct ipoib_mcast, rb_node);
if (memcmp(iter->mgid.raw, mcast->mcmember.mgid.raw,
sizeof (union ib_gid)) < 0) {
iter->mgid = mcast->mcmember.mgid;
iter->created = mcast->created;
iter->queuelen = skb_queue_len(&mcast->pkt_queue);
iter->complete = !!mcast->ah;
iter->send_only = !!(mcast->flags & (1 << IPOIB_MCAST_FLAG_SENDONLY));
ret = 0;
break;
}
n = rb_next(n);
}
spin_unlock_irq(&priv->lock);
return ret;
}
void ipoib_mcast_iter_read(struct ipoib_mcast_iter *iter,
union ib_gid *mgid,
unsigned long *created,
unsigned int *queuelen,
unsigned int *complete,
unsigned int *send_only)
{
*mgid = iter->mgid;
*created = iter->created;
*queuelen = iter->queuelen;
*complete = iter->complete;
*send_only = iter->send_only;
}
[IPoIB] don't compile debug code if debugging isn't enabled Don't build ipoib_mcast_iter_ functions if CONFIG_INFINIBAND_IPOIB_DEBUG is not enabled -- their only callers will not be built either. Also move the prototype for ipoib_open() to ipoib.h to fix a sparse warning. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2005-11-03 12:51:01 +08:00
#endif /* CONFIG_INFINIBAND_IPOIB_DEBUG */
Reference in New Issue Copy Permalink